Parliament and the People : How digital technologies are shaping democratic information flow in Sub-Saharan Africa

Parliament and the People : How digital technologies are shaping democratic information flow in Sub-Saharan Africa

mySociety Research - Rebecca Rumbul, Gemma Moulder, Alex Parsons

Executive summary

The publication and dissemination of parliamentary information in developed countries has been shown to improve citizen engagement in governance and reduce the distance between the representative and the represented. Over the last 20 years, there have been increasing efforts to use digital tools to facilitate this process, which has resulted in highly successful parliamentary monitoring websites such as TheyWorkForYou.com in the UK and GovTrack in the USA. Similar websites are now operating in Sub-Saharan Africa, among them Mzalendo in Kenya, People's Assembly and its sister site PMG in South Africa, ParliamentWatch in Uganda, and ShineYourEye in Nigeria. In addition to dedicated digital parliamentary information tools, a range of other digital channels are now popularly used to create and disseminate information about parliaments and politicians, most influentially social media platforms and online news media websites.

While it is clear that these channels are being used, it is not clear how they are being used, or why some digital tools wield greater reach or influence than others. Study into this area is in a nascent state even in developed countries, and while digital tools for democratic engagement are being implemented in Sub-Saharan Africa at pace, there is a paucity of research into their effectiveness and towards an understanding of the landscapes in which these kinds of digital tools thrive. Organisations working in this sphere should be asking tough questions about how the benefit of tools can be maximised, how failures can be informative for future projects, and how organisations doing good work can be supported to thrive in environments affected by instability and a scarcity of resource. Digital tools for democracy are no longer in their infancy, and organisations supporting and implementing them should now be taking an active, rather than passive, role in their future.

With the support of the Indigo Trust, mySociety has undertaken research to better understand how digital tools for parliamentary openness and engagement are operating in Sub-Saharan Africa, and how future tools can be better designed and targeted to achieve greater social impact. As one of the world's first civic tech organisations, dedicated to building impactful open source technology to empower citizens to hold their public institutions to account, mySociety has a unique history of developing digital solutions for civil society groups and governance organisations all over the world, and was well placed to conduct this research at the nexus of digital technology and information for government accountability.

The intended audience for this report spans civic technologists, development funding and philanthropic bodies, academics and any practitioners planning on building digital tools for democratic engagement in Sub-Saharan Africa. It is hoped that the findings of this report will inform future funding policy and help to shape how new digital interventions in the region are designed, implemented and maintained.

The report provides an analysis of the data and digital landscapes of four case study countries in Sub-Saharan Africa (Kenya, Nigeria, South Africa and Uganda), and interrogates how digital channels are being used in those countries to create and disseminate information on parliamentary activity. It examines the existing academic and practitioner literature in this field, compares and contrasts the landscape in each case study country, and provides a thematic overview of common and relevant factors in the operation of digital platforms for democratic engagement in parliamentary activity.

The report identifies a number of factors significant in the relative success or failure of digital tools in the case study countries, specifically concerning the quality and accessibility of data, the necessary socio-political environment to foster digital tool use, the importance of political buy-in, and the relevant expertise in building and managing organisations to run such tools.

The report makes the following recommendations:

  1. Conduct thorough scoping exercises in-country before committing to fund, build or implement a specific solution, and use the intelligence gathered to inform the final product.
  2. Work with in-country partners that have a good working relationship with their parliament, and ensure the digital tool is integrated into both their regular work and future discussions with parliament about improving civic engagement
  3. Make peace with solutions that aren't necessarily replicable, because a good digital platform that is built to be specifically appropriate to each country's unique governance structure will likely be better used and have greater longevity than platform structures replicated wholesale from other jurisdictions.
  4. Ensure that comprehensive, good quality, data sources are identified before trying to build anything, because poor or inconsistent data is one of the most common issues that threatens the operability of digital tools for parliamentary monitoring.
  5. Ensure ongoing, stable funding for maintenance and growth, and ensure this encompasses both development and non-development work, as without this, the platform will rapidly become out of date, and is likely to fall into obsolescence.
  6. Integrate digital tools as much as possible with relevant social media platforms, as shareable and user-friendly content is likely to be disseminated much more widely through these channels, than through visits to the tool itself.

The digital development landscape in Sub-Saharan Africa is an energetic and enthusiastic sphere of activity, and currently the 'youth bulge' is pushing for greater digital interaction and improved transparency and accountability of government. Digital tools to facilitate information flow between parliaments and the people are demonstrated in this report to be a valued, if currently under-utilised, resource, that have the potential to have significantly wider impact through so-called 'infomediaries' who are able to repackage information into a more consumer-friendly format. Using the recommendations in this report, it is hoped that future tools can be more effective and impactful in enhancing how parliamentarians and citizens work together for a brighter future.

Introduction

The nature of communication between citizens and their institutions, officials, and politicians has evolved significantly with the advent of the internet and corresponding social and mobile media channels, to a state that would have seemed impossible (and not necessarily wholly desirable) only 30 years ago. Whereas correspondence via letter or in person through a constituency surgery or other formal channel would have been the only way for the general citizenry to make direct contact with their officials in the past, multiple digital channels now exist to communicate information back and forth directly and at scale. Digital communication today also allows for anyone, anywhere, to create and mass-distribute political content, in a way that was traditionally only accessible to professional journalists and broadcasters. No formal qualifications, quality control processes or consideration of ethical issues are now needed for individuals or groups to author and disseminate information as 'news'.

The rapidity with which new space for political and official discourse has mushroomed online has outpaced the abilities of lawmakers to regulate the space, the ability of behavioural experts to understand how this form of discourse affects citizen attitudes and actions, and the capability of the tech giants themselves to identify and prevent pernicious uses of their own platforms. Society is only now in the early stages of understanding how the communication of official (and unofficial) political and parliamentary information is shaping societies and institutions, and what the implications of this shift may be.

Understanding these emerging issues is extremely complex, and landmark data manipulation scandals such as Cambridge Analytica's supposed influence on Brexit or the USA's allegation of Russian digital influence in the 2016 US Presidential election, have demonstrated not only the new ways in which political and official information can be manipulated and disseminated, but how incredibly targeted according to context that information is becoming. These stories illuminate the importance of subtlety, context and social ties in introducing and disseminating information relating to politics and government, and show that communication of information is necessarily different across different nations, identities, cultures, demographics and environments.

Much contemporary study has been focused on the implications of information dissemination around politics and policy in developed countries in Europe and North America; however, very few studies have examined the parallel phenomenon in Sub-Saharan Africa. Given the significant investments in digital and telecommunications infrastructure in African countries such as Kenya and Nigeria, and the large sums spent from various international development programmes and philanthropic organisations on fostering citizens' digital skills and improving governance, it seems prudent to begin to examine how political and official information is being created and disseminated in Sub-Saharan Africa. Connectivity is growing at a rapid rate across the continent: over 20% of individuals in Sub-Saharan Africa had reliable internet access in 2016 (World Bank, 2018), with this number projected to have reached 35% by the close of 2018. A 2017 GeoPoll survey demonstrated that in Kenya, Nigeria and Uganda, social media has become the primary news source for the younger 'millennial' age groups aged between 18-35, the largest age group on the continent, and while the concept of 'fake news' and unreliable media is now widely understood, it remains difficult for individuals to correctly identify genuine information shared online.

This report seeks to delve further into the information creation and consumption activities of institutions and citizens in Sub-Saharan Africa. Whereas much existing literature focuses solely on social media and the fake news phenomenon, this study examines information creation at an institutional level, and the transmission of that information downwards, horizontally and upwards between parliaments, media outlets and citizens, which involves social media as well as other media. In particular, this study looks at 'civic tech'; forms of technology that are developed outside of government, within the nonprofit and development sectors, and with an aim to empower and enable citizens to hold their governments and institutions to account, in one way or another.

The report has three main purposes:

In achieving these aims, the goal is that future projects and programmes designed to improve governance, democratic engagement and accountability in Sub-Saharan Africa will become more tailored and appropriate, and be of greater benefit to citizens and officials.

Problem statement

Significant funding for the development of digital tools and platforms has been funneled into projects in Sub-Saharan Africa over the last ten years, primarily by the international development and philanthropic community seeking to improve governance, democratic engagement, well-being, health, economies, agriculture and aid/relief programmes at scale and through low-cost means. The explosion of internet and smartphone usage provided a potentially fertile landscape on which to build better structures, services and empowerment tools for ordinary citizens, which was identified as one in which citizens could organise, mobilise and challenge their institutions effectively with minimal personal inconvenience.

As a result of development and philanthropic funding, a number of apps, websites and digital platforms such as Ushahidi, MapKibera, and ForestLink have been developed over the last ten years specifically to address barriers and challenges faced by citizens in Sub-Saharan countries. Many less visible tools have, however, proved unsustainable or unfit for purpose. Projects that had attracted high levels of funding, for instance the USEDATA project funded by UK aid and implemented by Montrose, found themselves floundering; apps that cost significant amounts went unused by their target user groups, and flashy websites quickly went out of date. While tech and international development gatherings around the world were quick to promote and praise innovation and concept, it was rare to hear about the numerous digital projects that fell by the wayside.

This severe lack of discussion concerning civic tech failure in Sub-Saharan Africa lies not only at the feet of those designing and implementing projects, but also those funding them. The replication of similar project ideas with similarly disappointing results has arguably continued far beyond what was necessary had technologists and development experts had a more effective and safer forum in which to discuss issues, rather than successes. As it is, many project implementers feared that an admission of failure would automatically reflect poorly on them as an organisation and result in a cessation of funding.

Between 2005 and 2015 digital tools and apps for democratic participation operating in developed countries were demonstrating success in engaging and empowering citizens. In the subsequent rush to replicate these successful digital projects in developing countries as a quick and economic democratic fix to corruption and political instability, it has increasingly become clear to technologists, funders and researchers alike, that the underlying problem has been that tech was assumed to be completely transplantable and scalable across very different countries, with no regard for the cultural assumptions that went into building it or the cultural traditions into which it was being implemented. While, at its core, the civic tech being aspired to consisted of little more than new conduits for information flow (and the ability to search and manipulate that data) between citizen and state, the way in which the tech was delivering this facility was clearly problematic across many of the tools in practice.

The overarching research question this study sought to answer, was therefore:

How do national and sub-national conditions in Sub-Saharan Africa affect how communications between citizens and governance institutions are conducted digitally?

This high level question is broken down into the following sub-questions in order to focus the research and encompass a range of relevant areas for examination:

  1. In what ways are parliaments creating data/information and making it available?
  2. What political and official will exists to engage in, and to improve, a two-way information flow between citizens and institutions?
  3. How are citizens consuming, contributing to, and sharing information on politics and parliamentary institutions?
  4. How does the accessibility, connectivity and quality of the data landscape shape how individuals and institutions communicate?
  5. What other as yet unidentified factors are shaping information exchange and communication between citizens and parliamentary institutions?

This study sought to address these sub-questions as a route to understanding the overarching research question, and used them to shape the interview discussions with participants and stakeholders. The international development community that focuses on improving governance, democracy and citizen engagement through digital development tends to hold a theory of change that is based on the openness and accessibility of information available to citizens and civil society being used by these constituencies to hold politicians and institutions to account. This theory of change also influenced how the study was framed, through the lens of change and improvement, rather than as a static picture of current activity. The research question and the information exchange landscape were examined in terms of data and information, and while specific tools and platforms were investigated for their effectiveness, the main focus of the study was on the meta-level, seeking to understand how the social, cultural and economic ecosystems in which digital tools are placed affect their success or failure. Using the information available on various digital tools employed in Sub-Saharan Africa as a starting point, the study was guided by considerations of quality, accessibility, resource, political and social will, and skills.

The resulting conclusions will re-examine these research questions and identify key areas in which environments can facilitate or frustrate effective digital communication and engagement between citizen and state.

Research methods

To best understand the processes of information flow between citizens and states, a qualitative approach was considered the most effective method. While a statistical review and assessment of internet usage and related communications and information processes would be beneficial in complementing the research, it was felt that this method of quantitative study alone, performed at distance, could not uncover the nuances hypothesised.

The qualitative method selected was a comparative case study approach, in which the information flow between institutions and citizens in four countries in Sub-Saharan Africa would be examined. A comparative element was considered necessary in order to highlight both similarities and anticipated differences in the forms of communication in Sub-Saharan Africa. Mindful of the significant cultural diversity across the region, it was not expected that actors in each country would exhibit the same behaviours or experience the same issues. As with any research project, budget constraints existed; however, it was felt that examining only two or three countries would provide too few generalisable conclusions, and therefore the decision to examine four was taken.

The four countries selected for study were Kenya, Nigeria, South Africa and Uganda. The criteria on which these were selected comprised:

In reviewing potential countries to include in the study, it became apparent that while digital tools had been developed in a large number of Sub-Saharan African countries, some were no longer in use, or usage was restricted to only limited areas around urban centres. Kenya, Nigeria, South Africa and Uganda were clearly the most suitable based upon the criteria set out above.

Research methods primarily centred around gathering information directly from individuals and organisations. The research was divided into three parts:

  1. Desk-based research literature review;
  2. Phone/Skype interviews with key stakeholders in each country;
  3. Case studies of four countries, including field visits and face-to-face interviews.

An initial desk-based literature review was conducted to examine existing knowledge of information flows between citizens and states, and to identify gaps. This encompassed published, peer-reviewed research in academic journals, as well as so-called 'grey' literature, which is produced by stakeholders, civil society groups and public sector sources. This review served to inform the research as a whole, refine the research questions, aid in creating the criteria on which to base the selection of case study countries, and inform the development of the interview questions to be posed to research participants.

In addition to the desk-based research, a small number of preliminary interviews were conducted with known individuals working on digital development and information flow between citizens and states. These were informal, scene-setting conversations, generally via Skype, conducted to produce a more detailed working knowledge of the sectors in each country and identify key areas of potential focus for this study. The interviews were not specifically structured in order to cover a wide range of potential interests, and were used to further inform the interview questions and identify research participants to interview.

Field trips to each of the four case study countries were conducted during 2018, primarily for the purposes of holding semi-structured interviews. A total of 65 interviews were conducted. Individuals interviewed included government officials and politicians, civil society groups working on democracy issues such as parliamentary monitoring and access to information, campaigners and community groups, journalists and broadcasters, specific interest groups, and digital development groups. The research participants interviewed were identified through existing contacts with organisations in-country, the preliminary desk-based research, recommendations, and snowball sampling. All interviews were conducted in English. While several individuals were content to participate candidly in these interviews in an official capacity, a significant number of interviewees requested anonymity due to personal security concerns. In order to protect the identity of those individuals, the names of the research participants will not be published. However, the majority have agreed to allow publication of their organisational affiliation to provide as clear a picture as possible of the rich source of information this report is based upon: this list can be found at the end of this document.

The questions posed in the semi-structured interviews were formulated according to the interviewee's affiliation, which was broken down into three main categories:

While many of the interview questions were similar across these three groups, categorising them in this way allowed for greater nuance. In all three categories, questions focused on the following themes:

Interviews lasted between one and two hours to allow for in-depth exploration of the themes of the research, and were conducted primarily in person, with a minority conducted using Skype where diaries would not allow for a face to face appointment.

Detailed notes were taken during interviews, and later coded manually to identify overarching themes and sub-themes.

Literature review

Nation state parliaments form the central core of government and legislative power, and are intended to represent the diverse views and needs of a country's citizens. As the means through which democratic government is made possible, political representation within a parliament or legislature is a necessity in a pluralistic society (Weymans, 2005), and is a fundamental feature of any functioning democratic structure (Dahl, 1971; Scully & Farrell, 2001). Political representatives are installed to represent the interests of a given constituency, bringing diverse and divergent views into the policymaking field (Scully & Farrell, 2001), and as such, politicians are expected to understand those interests, and interact with those that they represent (Alcoff, 1991; Goodin, 2004; Brown, 2006). Across the world, however, the practice of representative democracy manifests in different ways from the theory. The representative relationship in both young and older democracies is currently considered to be in disrepair (Foa & Mounk, 2016), and the gulf between representatives and the represented has come into sharp focus as a result of better education, new technology, and globalisation (Coleman & Blumler, 2009).

In a number of countries in Sub-Saharan Africa, parliamentary bodies remain oblique and distant from citizens, and institutional structure and governance has been historically weak (Edigheji, 2006). Authors such as Edigheji (2006) and Azevedo-Harman (2012), have pointed to a lack of vertical and horizontal engagement between institution and citizen as one of the key causes, and have noted that communication links between state and citizens are weak, with few meaningful opportunities for citizens to influence and engage with their representatives (Oni, 2013; Awah, 2013). While citizens are noted to be very active at the ballot box, even in countries with low levels of electoral integrity (Norris, Wynter & Cameron, 2018) deeper understanding of legislative institutions is low, and most citizens lack knowledge of how to demand accountability from their representatives, and in the culture of respect, lack the confidence or perceived authority to do so (Azevedo-Harman, 2011; Mattes & Mozaffar, 2012; Oni et al, 2016).

Parliamentarians themselves have also been shown to lack the knowledge and understanding to change these behaviours from the inside. Oni (2013) notes that African legislators have relatively low levels of expertise in political economy and historical norms, and do not regularly perform the constituency and outreach work done in more developed legislatures. Even the physical orientation of parliaments in Sub-Saharan Africa can sometimes be a deterrent to citizen engagement, as many parliamentary buildings and facilities have security levels preventing citizens from entering (Gberevbie, 2014). As a result, government remains distant from citizens, and policy and practice rarely reflects the public's high priority concerns (Veit et al, 2008). This enduring distance between citizen and state is understandably manifested in public mistrust of political institutions, and has been cited as a cause for decline in citizens' general confidence in their governments and related public institutions (Pantoja and Segura, 2003; Rosenthal, 2009). Digital technology has therefore provided citizens and civil society groups with new hope in changing these established norms.

In line with the global increase in internet accessibility and use, there have been corresponding developments in the use of digital technologies and ICTs within parliaments in Sub-Saharan Africa. The majority of national parliaments in the region have a functioning website with basic information on members, activity and access; however, few have information published and displayed in formats that are easily navigable to individuals not familiar with the institutional structure or language (Oni et al, 2016). Bwalya et al (2012) noted that the majority of African parliaments lacked the quality and volume of expertise necessary to implement good ICTs, which manifested as a resistance to the adoption of ICT within a legislative setting.

There is evidence of demand for digital tools to better link parliaments and citizens in Sub-Saharan Africa. An Inter-Parliamentary Union Hansard report (2010) noted that East African respondents to a survey, when asked what digital tools they would like to see MPs use to improve communication, requested digital tools that would facilitate direct engagement and communication between parliamentarians and citizens, and that represented a move away from traditional, centralised 'broadcasting' of information from the centre out (Robert & Namusonge, 2015). At present, parliamentary websites in the region remain at the 'information broadcasting' stage (Oni et al, 2016).

This study examines multiple directions of information flow, from parliaments downwards, from citizens to parliaments, through infomediaries, and horizontally between citizens, accommodating a whole spectrum of digital reflexivity in which messages are constructed, deconstructed and reconstructed as information passes through multiple hands (Lengel & Newsom, 2014). A difficulty with this is the way in which information has been historically used by governments, in particular in Sub-Saharan African countries, where information on governance has been crafted into 'usable histories and narratives' by parties and governments either taking or maintaining a grip on power (Pickover, 2014). This tendency to cherry-pick information that supports the ruling narrative has seen parliaments shy away from wholesale record-keeping and the presentation of neutral facts, in favour of recording and publishing only that information which reinforces existing power structures (Pickover, 2014). Parliaments as institutions in their own right are therefore not currently migrating to strategic organisational use of ICT or digital, and as a result, are not producing the necessary data or information needed by citizens, civil society and infomediaries to enable them to hold government to account.

While parliaments as institutions in Sub-Saharan Africa are not generally using social media as an information tool, individual politicians and parties are shown within the existing literature to be slightly more connected. However, the current evidence shows that this digital engagement tends to be almost entirely confined to election periods. Citizens are using social media tools to talk about politics, and there is evidence that this usage spikes during election periods, as it did in Nigeria, where Twitter use increased tenfold during the last election period (Bartlett et al, 2015).

In Nigeria, authors such as Ajaye and Adeoste (2015) and Bonin (2013) have discussed how the population is migrating towards consuming political and journalistic content via social media channels rather than directly from media outlets, and as such, citizens are demanding more information online than ever before. In Kenya, the presidential elections of 2013 are remembered as the country's first digital campaign, with the candidates Uhuru and Ruto running their operation with acknowledgement that social media platforms would play the same role as traditional media (Robert & Namusonge, 2015). Similarly, in Uganda on June 14th 2015 the former Prime Minister John Patrick Amama Mbabazi embraced social media, announcing his bid to run for the presidency on YouTube in response to alleged insider information warning that state operatives had planned to block him from making the announcement at a formal press conference. He would later follow this up with an online press conference on July 6th 2015, the first of its kind in Uganda, offering any individual on social media the opportunity to ask him any question. This session, using the hashtag #AskAmama, was held at his home, and lasted for over 3 hours (Kamp, Messerschmidt & Rugambwa, 2016). The electoral focus is slightly lower in South Africa, where a number of high profile social media campaigns have influenced wider policy and decision-making outside of the electoral process. Prominent student campaigns include the #FeesMustFall campaign to pressure the government into lowering fees, and the #RhodesMustFall campaign in Cape Town demanding the removal of statues depicting the controversial figure (Bosch, 2017).

In addition to social media platforms and the official websites and digital channels run by parliaments, a range of digital platforms run by civil society groups are in operation across Sub-Saharan Africa that in one way or another enable citizens to hold their public institutions to account. Parliamentary monitoring sites run by civil society groups with the express aim of enabling citizens to see what is being said in Parliament and how parliamentarians are voting are in operation in South Africa, Kenya, Ghana, Zimbabwe and Nigeria, and the limited research conducted into these sites shows that while citizens value their existence, user numbers are relatively low (Rumbul, 2016). Other sites operated by civil society groups with similar principles of improving governance and engagement have also demonstrated lower than hoped-for user numbers, and have suffered from issues with implementation and buy-in from the political and institutional class (Anderson, Hudson & Feruglio, 2018). Open data has been shown in a number of studies to be insufficient to catalyse good digital tool development and better governance in and of itself (Anderson, Hudson & Feruglio, 2018), and even where data is available, there remain cultural and social barriers to holding government to account that often are not surmounted even where the data and tools are in place (Roberts and Hernandez 2017).

While use of digital tools and social networking platforms by parliamentarians has been shown to be patchy, there is a clear demand from citizens for their representatives to be accessible through these channels (Font, Wojcieszak & Navarro, 2015). Investigating the use of social networks by members of the East Africa Legislative Assembly, Robert and Namusonge (2015) found that only 73% of parliamentarians sampled were registered users of these sites, and used the services for their own leisure, rather than assuming any obligation to engage with or respond to citizens. This is important when considering how connected citizens feel to their representatives, because this feeling of connection also establishes understanding of the political system and confidence in politicians. A survey of constituents in the UK showed that citizens that had merely visited their MP's website felt three times more connected to their MP than those who had had no contact with them at all (Coleman & Blumler, 2009). Previous research by mySociety (2015) has shown that individuals who use parliamentary monitoring sites believe that the very publication of information on parliamentary activity compels representatives to act differently from how they would act if that information was concealed. Over 90% of respondents to that research in Kenya and South Africa felt that being able to see parliamentary information through digital platforms enabled them, at least in some small way, to hold their governments to account. Whether on social media or on another online platform, therefore, citizens are significantly more likely to have trust in their government when they are able to see what that government is actually doing.

This brief literature review has demonstrated that there is use and appetite for digital information flows between citizens and parliaments, but that this is currently not happening in a structured or easily accessible way, and that governments and parliamentarians remain too free to engage through these channels only when it suits them. There is little literature on how other digital tools are being used for democratic accountability in Sub-Saharan Africa, and a paucity of information on how citizens themselves feel about using these kinds of digital platforms. The interview questions used in this research are in part based on the themes raised here, and aim to begin to fill the significant gaps in knowledge that the literature review has exposed.

Case studies

The findings of this report will, for the most part, be presented thematically in the next section. However, this section provides a brief overview of the four countries in which the case studies took place. The purpose is not to provide an exhaustive analysis of current political, social and economic activity in each country; rather, it is to tease out the specific contextual factors that are unique to each country and explore the implications these may have in shaping how citizens and institutions are communicating.

KenyaSouth AfricaUgandaNigeria

Population (2018 predicted)

48 mil

57.4 mil

38.8 mil

193 mil

% Population in urban areas

0.270.660.230.5

Mobile coverage

0.89080.99910.9939

Individuals using internet, % (2015)

0.4340.490.17710.4262

Active mobile-broadband subscriptions per 100 inhabitants (2015)

15.559.52118.3

Mean download speed (MBps) (2018)

10.116.382.391.86

Average daily cost per GB per day (figures adjusted by cost of living and standardised against Kenya)

14.10.91.9

Source: World Bank Statistics, cable.co.uk broadband speeds

Data Pricing Comparison

To determine the relative cost of data on daily plans in the different countries, the data plans of three or four telecoms companies in each country were examined (September 2018).

CountryProviders

Kenya

Airtel, Faiba4G, Telkom

Uganda

Airtel,MTN, Smile

South Africa

CellC, MTN, Telkom, Vodacom

Nigeria

MTN, Airtel, Glo, 9mobile

Providers offer a variety of packages covering different time ranges (commonly daily, weekly and monthly) with different amounts of data included within this limited time period. This report compares the average cost of data packages covering one 24hr period instead of per GB, because the cost of individual GB units varies widely across packages. Daily packages are therefore the most effective unit across which to compare the four case study countries. In South Africa several providers offered hourly data plans which were excluded from comparison — although their existence is an indication that the relatively high cost of daily data in South Africa has created a market that provides cheaper (but worse value for money) data packages. Packages that were aimed at evening hours or could only be used on weekends were excluded.

Different costs were converted into USD to calculate for each plan the '$/GB' and the '$/GB/day'.

The minimum, average and maximum $/GB/day are shown below. In brackets is the cost relative to Kenya adjusting by the Numbeo Cost of Living Index (which changes the scale of the differences, but not the ordering).

For daily plans:

CountryMinimumAverageMaximum

Kenya

$0.49 [1]

$2.54 [1]

$6.20 [1]

Nigeria

$2.74 [4.7]

$5.74 [1.9]

$13.71 [1.9]

South Africa

$3.52 [7.9]

$9.59 [4.1]

$17.58 [3.1]

Uganda

$1.31 [2]

$3.15 [0.9]

$5.23 [0.6]

For all plans:

CountryMinimumAverageMaximum

Kenya

$0.004 [1]

$0.79 [1]

$6.20 [1]

Nigeria

$0.0069 [1.5]

$1.18 [1.3]

$13.71 [1.9]

South Africa

$0.0019 [0.5]

$1.7 [2.4]

$17.58 [3.1]

Uganda

$0.0036 [0.7]

$0.81 [0.8]

$5.23 [0.6]

These tables show that the cheapest plans show the greatest disparity between the countries, with the cheapest plan available in South Africa being 7.9 times more expensive in GB/$/day (adjusted by cost of living index) than the equivalent in Kenya. Looking at the average of the daily plans, Kenya and Uganda display rough parity, while Nigeria was twice as expensive and South Africa was four times as expensive.

Examination of the costs and relative pricing for all plans (including weekly, monthly and longer plans) displays far greater parity of pricing for those who can afford longer plans, while the best value plans show variance within 50% of the price in Kenya. For the average of all plans, South Africa is a clear outlier with data costing 2.4 times that of Kenya, while Nigeria is only 30% more and Uganda 20% less.

Kenya

Kenya is one of the frontrunners in Sub-Saharan Africa when it comes to recognising the importance of tech and digital and taking steps to foster their development, both in-country and internationally. The launch of M-PESA in 2007, a technology developed in Kenya specifically for a population with little access to traditional banking, was a game-changer for many. Soon after, the post-election violence experienced in Kenya during 2007–2008 catalysed further innovation. A group of concerned tech entrepreneurs started collecting reports of violence and disruption, through email and SMS, and uploading them to Google Maps. This was the birth of Ushahidi, a versatile information-gathering and interactive mapping tool now in use in multiple places around the world. These early successes ingrained in the Kenyan population an appreciation of the efficiencies of technology in everyday life, and the corresponding structural investment by government and private sector forces in infrastructure for mobile data and reliable connectivity is ensuring a fertile landscape for future development.

The implementation of a five-point policy on digital development became a key driver in the paradigm shift in Kenya's economic programme. It focused on the leveraging of ubiquitous mobile platforms in order to build applications, the development of reliable ICT infrastructure, the creation of locally made content, the growth of human resource capacity, the development of public-private partnerships (in particular, with the mobile and banking sectors), and the creation of employment opportunities for the mushrooming youth population entering the workforce. Developed from this strategy was the Kenya Open Data initiative, which sought to provide as much open data as possible to entrepreneurs in order to facilitate the development of the kinds of digital tools the market needed. These initiatives, along with the addition of the TEAMS cable (the first fibre-optic cable laid off the Eastern Seaboard of Africa), were crucial steps, and heralded the age of cheaper telecommunication access in Kenya. As such, new opportunities were created, and programmes to subsidise broadband for all universities and create hubs like those in Nairobi's now famous Ngong Road area were implemented.

Significantly for Kenyans, the political and economic importance of connectivity has grown to such an extent that illegitimate political tactics to disrupt the internet as a form of political communication are no longer deemed acceptable. Whereas a number of Sub-Saharan countries have, in the past few years, seen internet blackouts used by governments to disrupt elections or democratic activities, this is now considered a step too far in Kenya. During the contested election of 2017 in Kenya, many observers and activists were anticipating an internet blackout that never materialised, and official sources interviewed for this study admitted that such crude tactics were no longer an option. One participant noted that:

The government needs the internet just as much as the opposition does. It does not benefit the government to close down the internet. And . . . the financial implications of turning off the internet mean that all those people in government who are from families that own the banking and communications companies would be losing money. It can't be done.

Primarily, the political elite hold far too many business interests in finance and telecoms that rely on internet connectivity to believe that blocking access to the internet is a viable option. If internet connectivity is blocked, then financial transactions and the millions pouring into the telecoms companies' coffers every second take a hit, too. No one wants to sacrifice profit and business confidence for what is likely to be only a marginal effect on voting behaviours. The success that has been demonstrated in the investment and development of the digital environment in Kenya has had a significant impact upon the way in which politics and political activity is now conducted. The economic and political argument for keeping digital channels open at all times now triumphs over the view that a short internet blackout can help to consolidate electoral victory. There is, however, still political meddling when it comes to telecommunications in Kenya: it just uses more subtle approaches. One participant in this study said their project using SMS short codes to engage with and collect citizens' views on current political topics was affected due to telecoms companies being told not to send out messages that were political during the 2017 election.

The digital environment within government may be less advanced than across the country as a whole; nonetheless, the Kenyan parliament and government are more digitally advanced than their counterparts in neighbouring countries. Open data concerning parliamentary activities and politicians is available and is used online in tools such as Mzalendo; and a strong civil society is able to engage directly with ministers and officials on a range of issues. Culturally speaking, Kenyans are self-described "political animals" as several of the research participants put it, and are, compared to their geographical neighbours, more proactive in engaging directly with politicians and officials online and in person.

Ethnicity plays a significant role in politics in Kenya, and those that participated in this project were in agreement that it is impossible to talk about politics in Kenya without also talking about ethnicity. That said, there was no indication that different ethnicities used digital technologies or were engaged in politics or parliamentary interests in different ways. Both digital literacy and interest in parliament were considered to be at a broadly similar level across ethnic groups, and are therefore treated as such in this study.

As a digital leader in the East Africa region, and one of the largest digital services economies in Africa, Kenya is therefore a fascinating and unique case to study on these themes of digital development and communication.

Nigeria

With a population of over 190 million, and with mobile penetration currently at approximately 85%, Nigeria is the largest mobile market in Africa and one of the continent's hotspots for smartphone usage due to comparatively cheap access to handsets. While the digital infrastructure in the country as a whole remains in need of significant development, the country's access to undersea cables ensures that the capacity can be realised as soon as the corresponding domestic work can be carried out. Nigeria's ICT strategy and emerging partnership projects with DFID, the EU and investment by large international digital and domestic private sector companies to improve the reliability of the internet are creating a solid foundation for progress. The government has actively attempted to prevent a rural/urban divide from emerging with the creation of the National Information Technology Development Agency (NITDA), tasked with bringing digital literacy to more remote regions of the country where internet penetration is at its lowest. This agency has created IT Innovation Hubs, which are focused on improving digital literacy amongst rural groups, but also amongst groups from low skilled or low paid occupations, the unemployed, and those with disabilities. As such, the majority of Nigerians are well positioned and motivated to create and consume information digitally.

Despite the work of NITDA, Nigeria's main urban centre Lagos is currently significantly more digitally developed than other parts of the country, and the digital and communications focus here tends towards entrepreneurship and e-commerce. While there are a number of civil society groups operating effectively out of Lagos, the distance from parliament and government means that many of the issues addressed are more locally focused around the city and the unique needs of its citizens. The capital Abuja, in the geographic centre of the country, hosts a thriving civil society, but one that is quite loose and unable to take full advantage of digital technologies in engaging with government and parliament, due to the reluctance of official sources to produce and open up the necessary data. Civil society here is somewhat fragmented, with a range of organisations concerned with transparency, democracy and accountability operating not only independently of each other, but somewhat in isolation, and these organisations often find themselves in competition with each other for funding. This reduces the potential for effective civil society cooperation, and has resulted in the repetition and reinvention of products, services and activities, as organisations seek to work unilaterally and gain maximum funding, rather than pool resources towards common goals. Such behaviour is likely to reduce the effectiveness of civil society as a whole, and in an increasingly tighter funding environment, organisations are likely to face the prospect of either changing the way in which they operate, or ceasing to operate.

The parliament and government in Nigeria also operate at arm's length from much of civil society, with low levels of meaningful engagement and of political will to engage further. Within the civil service, the government has created a digital Government Service Portal and Government Contact Centre, deploying ICT to enhance transparency, efficiency, productivity and citizen engagement. These services are designed to create a single point of entry to federal government services, to enhance accountability and improve the delivery and quality of public services through technology-enabled civic engagement, to increase public administration efficiency, to improve user productivity by reducing the extent to which information is contained in silos, and to facilitate efficient response to citizens through the Government Contact Centres being set up in the six geopolitical zones of the country. At the time of writing, ten government processes from the Federal Ministries of Education, Health, Agriculture, Industry, Trade and Investments and Communication Technology have been automated.

While civil service and departmental migration to the use of ICT is evident in Nigeria, it is less so in the federal legislature. The institution as a whole engages minimally in the use of ICT to record and publish information, and there is no real culture of engaging collaboratively with civil society. The post of Minister for Civil Society within government was described to this study by a senior official as "a punishment post" and one "given to a parliamentarian as punishment for acting against the government on other issues". Whereas in the UK, for instance, voting against the government might see a ministerial MP banished to the backbenches, in Nigeria the rogue member is given a senior role dealing with a sector considered troublesome. This was of course a somewhat flippant comment, but one made by an individual intimately familiar with the role, and it illustrates the general sentiment of the Senate and Assembly towards working with civil society. A participant who had previously worked in the Senate administration told of an 'us versus them' attitude within the institution, where you were either 'inside' or 'outside', and if you were lucky enough to be 'inside', your loyalties would need to immediately switch to the party and your colleagues, rather than those on the 'outside'. Such priorities no doubt shape the way in which the Nigerian institutions create, disseminate and share information with citizens.

The Nigerian Senate has a low level of online presence as a public body, but has monitored the growth of wider digital communications online in detail, and in 2018 passed legislation concerning the criminalisation of hate speech online. This is somewhat controversial legislation given the scope of the legal terms used, and civil society groups have expressed concern that it may be used to stop dissent or criticism of the ruling party. At the time of writing the legislation had not been used to charge anyone with a crime; however, this piece of legislation demonstrates that parliamentarians in Nigeria are not deaf to digital communication, and are willing to legislate to shape the online environment they find themselves in.

As with many countries in Sub-Saharan Africa, corruption remains a significant barrier to transparent, accountable and responsive governance, and hampers the engagement of citizens in public and governance issues. The majority of research participants in Nigeria in this study noted that the lethargy with which the public generally approached politics and governance was a direct result of years of corruption, about which nothing meaningful ever seems to be done. Trust between government and citizen has mostly broken down. One participant explained that "People are tired of being angry about it. It is easier to look away". Because of this ambivalence, few people in Nigeria try to engage with their parliamentarians, and few feel they are able to challenge public institutions to meet their obligations. There is also a genuine reluctance to challenge a status quo in which citizens can be confident that they will not be worse off in a less corrupt system. Many people have known only corrupt systems, and have a reasonable fear that if those systems changed, they would be out of pocket or less secure than under the old system. Corrupt institutions drive citizens to act against their best interests, and as such, it is harder to identify where a good starting point for reform could be.

As explored briefly in this section, the data, civil society and governance landscape in Nigeria provided a rich and varied prospect for the study of information flow concerning public matters. As the largest economy and leading digital nation in West Africa, the findings of this research may not only benefit groups working in Nigeria, but in neighbouring countries too.

South Africa

As in the cases of Kenya and Nigeria, South Africa is a leader in digital technologies within its region of the continent. Internet penetration has been indicated to have reached 60% in 2018, and is still growing. While, as in many other countries, usage is concentrated within urban areas, the government is implementing an ambitious plan to broaden access, improve skills and increase economic engagement in digital technologies across the whole country. Part of this strategy includes consideration of the 'fourth revolution', a reference to concern about the massive effects of technology on the workforce and working practices of the future. Both national and municipal governments are expending resources on understanding how this digital shift will happen and what the implications are. Indeed, in pockets of the country, digital services are already as ubiquitous as they are in any global city, and have been encouraged to become so by both consumers and policy-makers seeking to make cities safer — because the rise of digital consumer services is aiming to cause a fall in petty and violent opportunity crime. Whereas in the past people would need to carry cash for taxis, drinks, restaurants, general shopping and myriad other things, digital services such as Uber and ApplePay now ensure that places such as Cape Town are becoming practically cashless societies that disincentivise petty theft simply by removing cash from the equation (IOLNEWS, 2018).

In addition to the attractive safety benefits of the migration of services online, these shifts in the economy are also creating employment, a key pledge of the South African government over the last decade. These new jobs are not only within the 'gig economy'; digital technologies have quickly been absorbed into private enterprise across the country over the last 15 years, and with them has grown demand for the associated highly skilled and technical workers necessary to support the field. South Africa's relative stability and unique history have endowed it with certain advantages in mobilising quickly to take advantage of the digital revolution; however, this could not have been achieved without significant government support.

While support for the growth in digital technology has undoubtedly been provided by government, digital expertise within government has not been developed at the same pace. Comparatively, this is not unusual, however very little will within government in South Africa exists to develop this expertise or to develop the institutions themselves into digitally accessible entities. While many politicians are online in one way or another, many are reluctant to use digital media as a form of communication, and many rural and older MPs lack digital skills. Some participants noted that while this could be considered a deficiency, it realistically represented the nation as a whole.

South African MPs, unlike those in the UK, Europe and USA, tend to more closely reflect the social and economic demographics of the nation (a democratic necessity following the fall of apartheid), and therefore it was expected that many sitting MPs would not have had expensive educations, access to high priced emerging tech or legislative experience. The parliamentary administration provides support to fill these gaps in knowledge; however, this can be problematic. One research participant in this study gave the example of the official legal counsel to parliament, who is viewed by committee members of all parties as 'their' advisor for the committee, when in fact this person represents the interests of the President, not the interests of the committee. Additionally, participants noted that parliamentary researchers for the committees have significant liberty to pursue issues of personal interest to them and potentially influence the way in which these issues are received and consumed by the committees. This is not to say that they do, and participants in this study have been both complimentary and admiring of the work that the researchers conduct. However, potential room for exploitation within the system nevertheless exists.

South Africa has a broad and engaged civil society that generally makes every attempt to engage with government and inform the legislative process. Government committees welcome a wide range of contributions, and occasionally tour the country to gather the views of those unable to attend committee sessions in Cape Town; however, while politicians are very aware of their obligation to accept input, they will not necessarily take it forward into policy formation. A number of research participants described the provision of evidence to government as 'tick-box exercises' rather than with any expectation that best practice would be achieved in their areas of expertise.

While such claims are disappointing, the majority of civil society groups are upbeat and energetic about South Africa developing towards a brighter future. Their frustrations with the operation of parliament are tempered by their appreciation of the way in which rights and freedoms are generally being upheld, in contrast to many of their more politically tumultuous neighbours. One participant noted that the right to protest peacefully was considered so sacred by government that it was actually becoming a serious nuisance to the residents and workers of Cape Town, who are inconvenienced by protests bringing the city traffic to a halt near parliament more often than they feel is tenable.

A vibrant section of South Africa's civil society encompasses a large and skilled civic technology and open data community eager to engage with public institutions in improving public services, and well networked both within South Africa and the global community of tech-for-good enthusiasts. Groups such as OpenUp in Cape Town and Open Data Durban are at the forefront of digital efforts to revolutionise how citizens and governments in South Africa interact, and work on multiple projects to improve public service provision. These groups are active internationally, and champion the re-use of open source tech to solve local issues digitally. These groups are also keenly aware of the pressing domestic need to avoid creating or exacerbating a digital divide within the country, and so have an emphasis on tech that can reduce the gap between the affluent and economically disadvantaged, and give those in lower income brackets the means to engage in public life on an equal footing.

On the one hand comparatively affluent and developed, while on the other suffering mass inequality and development challenges, South Africa is a unique and fascinating country in which to study official interactions between institutions and citizens. The political structures and equalities considerations here are vastly different from much of the rest of the continent; however, they provide interesting contrasts and unique perspectives on the same issues. These differences nonetheless pale in comparison to some of the fundamental factors in political and official communication, as the thematic analysis will demonstrate.

Uganda

Unlike South Africa, Nigeria and its own neighbour Kenya, Uganda is not a digital leader in its region, and is a comparatively geographically and economically small nation. Uganda has different development challenges, as well as different economic and structural issues to address, many of which cannot directly compare with the previous three case study countries in this study — but which may still provide valuable insights when considering development activities in countries with similar geopolitical considerations. Uganda is a landlocked nation with relatively limited diversity in its economy, that has suffered from a lengthy civil war and years of entrenched corruption overseen by a president unwilling to step down from office for over three decades.

Approximately 40% of Uganda's 40 million population frequently access the internet, and while connectivity is relatively good across the main populated areas of the country, it is less reliable in the rural north, where the recent civil war raged, and which now plays host to significant numbers of refugees fleeing neighbouring South Sudan. Connectivity pricing is structured differently in Uganda from in other countries in the region, which until recently made it more expensive to browse the web but cheaper to access specific social media sites. This has changed since the implementation of legislation to tax social media access in July 2018.

The taxation of social media aside, the Ugandan government has passed several pieces of legislation over the last 15 years concerning the use of ICTs, and has drafted a number of policies to foster the use and economic potential of digital technologies. However, many of these policies are fragmented and housed within departmental silos, preventing strategic cooperation and causing disruption and incoherence at the delivery level. Programmes to support the development of skills and incubate start-ups are again operating in theory, but in reality vulnerable to incoherence and underfunding.

The government of Uganda has also been more reluctant than others in the region to subscribe to measures to improve transparency, accountability and responsiveness, or to engage actively with the electorate and civil society. Uganda is the only country in this study that has not moved to become a member of the Open Government Partnership, and despite its natural resources industry, is not a member of EITI (although neither are Kenya or South Africa). Its recent introduction of taxation on social media usage is seen by many commentators and participants within this study as effectively a tax on social and political mobilisation, with the aim of curtailing criticism of the government and preventing public dissent. The tax is particularly pernicious, as the majority of low-income earners in the country use mobile deals that solely access the top three to five social media platforms, without access to the web in general. In a country where Facebook has been described simply as 'the internet', this tax will clearly reduce the ability of a significant number of citizens to use social media to share information or interact in any way with social or political online debate.

The government in Uganda is correspondingly reticent to engage too much with civil society and NGOs. The Ugandan government has, periodically over the last 15 years, come into outright conflict with the civil society sector, and legislation has stifled the field with complex regulation relating to public order and freedom of speech. While several participants in this study talked of a 'good' relationship with government and parliament, this was a relative term, with those relationships painstakingly cultivated over many years and operating with a high level of caution. Certain civil society groups felt able to provide advice to government, but only in a very gradual and respectful manner, and their work with government to make changes or recommendations to policy has in many cases depended on the individuals within government and their interests. Interactions with government were focused on engagement and ongoing dialogue rather than unilateral or confrontational approaches. The media also walks a fine line between reporting on government and observing the safety of its journalists. While state 'kidnappings' and imprisonment of journalists without due process have subsided over the last two years, such incidents still occur, and one participant in this study noted that "we only have one journalist at the moment who is being held by the state. This is an improvement on how it has been in the past". That said, the current political climate in Uganda has deteriorated in the months between the field work and the publication of this research, and no long-term stability is yet in sight.

In spite of precarious relationships with government, civil society has achieved a plurality and retains an enthusiasm for engaging with government to improve governance and increase transparency and accountability. Groups here work very collaboratively, and these strong links strengthen the sector and ensure its messages are coherent. Parliamentary monitoring operates in the form of ParliamentWatch.ug and their associated social media offering, and Freedom of Information is championed by both CIPESA and the Africa FOI Centre based in Kampala. While civil society has identified the potential for technology and digital to revitalise governance and citizen-government interaction, progress is frustrated by a lack of infrastructure, the cost of connectivity, the limited data available and the government's unwillingness to commit to progressive practices.

Uganda is a country of significant contrasts, where certain industries are extremely well developed, and others remain constrained by the current government's policies and practices. Its government and governance structure, as well as its economy and wider infrastructure, present a very different case study to Kenya, Nigeria and South Africa, but one that includes considerations relevant to other Sub-Saharan countries not included in this study. It brings to light important points for consideration when implementing new digital projects in similar environments.

Thematic findings

This section provides a comprehensive discussion and summary of the findings of the research. The findings and analyses are grouped thematically, rather than presented as four separate country case studies, in order to best draw together the common factors that shape how information and data are created and disseminated, and how citizens and governments interact digitally. As is the case throughout this research paper, the focus within each of the themes discussed is on how such findings can be used to inform better funding practice and development activity in order to leverage the greatest benefit for both citizens and their institutions.

Connectivity landscape

mySociety specialises in creating open source software and digital services for civil society groups and citizens to engage with politicians and institutions. As such, the organisation has,over the last 15 years, encountered numerous funding programmes calling for new digital products that would address a range of development issues. Rarely have tender or application documents discussed or made provision for a fact-finding, user-research or 'discovery' component to allow organisations carrying out the digital work to truly understand how the average user would be positioned to engage with the new platform being conceived. On the basis of the evidence gathered for this research study, this is one of the most fundamental errors made by those organisations attempting to work in this complex field. The tendency for agencies to arbitrarily group together countries geographically, or group 'priority countries' according to internal institutional definitions and indexes, somewhat deprives each country of its individuality, and conveys an assumption that all countries in that grouping are more or less similar in terms of their social, political and economic landscapes. Additionally, the idea that a technology developed in Europe or the USA, with all the cultural assumptions that are baked into it, can be parachuted into a country with a completely different culture and infrastructure also marginalises the fact that every country is fundamentally different.

At the most basic level, the way in which citizens and institutions access the digital environment is key to the success of any digital intervention. In developed countries, it is taken for granted that an average citizen will be using a smartphone, is also likely to have access to a desktop computer, tablet or laptop, will have a reasonable quality mobile signal or broadband, and will probably have a contract with a mobile network that provides more data and additional services than an average citizen will use per month. Not only is this not the case in the majority of Sub-Saharan nations, but this study found that in each country the data access environment differed significantly.

While the African continent largely leapfrogged the landline telephone and rapidly embraced mobile technology, the state of connectivity and the hardware required differs greatly from place to place. A majority of Africans are able to access some sort of mobile technology, but as yet, smartphone users are in the minority. Hybrid phones and 'dumb' phones remain the most widely used handsets in 2018.

The way in which individuals purchase and consume mobile technology also differs greatly. The creation of M-PESA, one of the most prevalent and revolutionary mobile facilities in Africa, is testament to this. M-PESA was developed following the observation of individuals settling small debts or loaning small amounts of money through the exchange of mobile phone credit. In countries where few people held bank accounts and where small change was terminally scarce, this enabled 'money' to change hands both amongst friends and family, and in the wider informal economy. This was a technology developed with not only a deep understanding of how individuals could use mobile technology, but how they were already using it to work around everyday obstacles.

Uganda is a key example of a country in which digital platforms for citizens to engage with government have failed to attract a critical mass of users. Leaving other factors aside for the moment, at the most basic level, the majority of mobile data users in Uganda were until recently using data plans that could access only the top three to five social media platforms - Facebook, Twitter, WhatsApp, and optionally Instagram and Snapchat. Cheap data plans for just these sites meant that, for the most part, individuals could engage with their wider networks in a cost effective way.

Because these sites are so significant in Uganda for everything from getting news to keeping up with old friends (indeed, Facebook is so ubiquitous that it essentially is the internet to many people), most did not feel the need to pay extra to browse general websites. One might build the most culturally sensitive and all-singing all-dancing web platform — but still it would not be able to overcome the fact that the majority of people in Uganda would be unlikely to use it because they do not have general web browsing access.

South Africa presents a different consideration, as here the cost of any data package is high in relation to average income. While there are pockets of South Africa with public and free Wi-Fi provision these tend to be in gentrified urban areas. Citizens who do not have regular access to Wi-Fi are generally very aware of their data usage and make attempts to minimise it to keep down costs. As such, people are actively weighing up the cost-benefit of visiting websites or using digital platforms and apps, and potentially judging that they cannot afford both to engage with an institution and to keep up with friends and family via WhatsApp.

The case is different in Kenya, where data is more popularly bought in bundles of GB or MB, often with a set time limit for usage. For as little as 0.49 USD an individual can browse up to 150MB in a 24hr period, and even smaller packages are available for browsing for as little as one hour. Data costs are similarly low in Nigeria, with 75MB available for approximately 0.25 USD, and smartphone usage is rising sharply. The data environment is therefore much more amenable to accessing the breadth of the internet, allowing individuals a cost-effective choice in exercising their browsing habits.

This means that standalone apps and general websites are likely to get much higher volumes of traffic in Kenya than in Uganda, and this is borne out by the usage numbers of sites such as Mzalendo, a Kenyan parliamentary monitoring site, in comparison to ParliamentWatch's similarly focused website in Uganda. Whereas Mzalendo receives extremely high traffic to its website — anywhere between 300,000 and 400,000 unique visitors per year — only modest numbers, approximately 45,000, pass through ParliamentWatch.ug per year. Both sites are comprehensive and informative, and are run by extremely educated and energetic groups, but there remains the fundamental difference in usage. The fault does not lie with the group managing and maintaining the site; rather, the project was built on assumptions that do not necessarily reflect the reality of data accessibility and web browsing behaviour in Uganda. While the ParliamentWatch website has relatively low visitor numbers, its Twitter feed, conversely, is immensely popular, with a following of more than 54,000 that puts Mzalendo's 14,000 in the shade — especially impressive for a comparatively small country. This group have circumvented the issue of low website usage by live-tweeting what is being said in Parliament to their followers.

These examples demonstrate that information on the discussions within, and machinations of, parliaments in Sub-Saharan Africa is of interest to the public, but that accessing this information requires different approaches dependent upon the data access environment in question. The high data costs in South Africa make citizens think twice before searching for anything, and therefore sites that require lower levels of data may be more popular. Despite sharing a border, the way in which citizens have thus far accessed parliamentary information in Kenya and Uganda is significantly different and to a certain extent dependent upon the data connectivity environment. These considerations form the very first hurdle in setting up new digital tools in Sub-Saharan Africa, and to ignore such issues is to limit the potential impact of a tool before it has even been implemented.

Parliamentary structures

As in many countries around the world, the levels of common understanding of governance structures among citizens are variable in Kenya, Nigeria, South Africa and Uganda. Interviewees working in policy, politics, journalism and civil society generally had a good grasp of the policy-making and legislative process, however, this diminished the further away an individual's employment related to these activities. Large swathes of the populations in these countries, particularly at the lower income and rural end of the spectrum, have very little understanding of how official governance processes work, and are thus more vulnerable to participating in corrupt practices in an effort to address their issues with government. Several participants in this research indicated that the ignorance of constituents was favourable to their representatives, either because they would not trouble them with additional work, or because these constituents would be open to paying a bribe in order to get their issue addressed, not knowing that the representative was under obligation to represent the voices and concerns of their citizens without such encouragement. This incentive to keep constituents in the dark about their rights manifested in different ways across the case study countries.

In Kenya, whilst the majority of citizens knew who their MP was, they were much less likely to know what an MP is actually supposed to do. The additional layers of devolved government, created following devolution of certain functions to a regional and local level, has also muddied the waters here. As such, citizens regularly fall back on their historic understanding of what their representatives are for, in many cases assuming that obligations include activities such as paying for school fees or attending funerals, rather than holding government to account for any failures to fulfil its obligations. In Nigeria, while individuals again often know who their representative is, citizens find these representatives to be increasingly unavailable after taking office, and therefore find they have no-one to approach about constituency issues. Members of the House of Representatives and the Senate also fail to return to their constituency seats with any frequency, preventing citizens that do not have a pre-existing relationship with them from cultivating one in person.

South Africa is different again, as a significant number of citizens have no knowledge of who their parliamentary representative is — primarily because the electoral system requires no geographic ties between representative and constituency. South Africa uses a proportional party list system to select representatives, which is significantly different from constituency-based, first-past-the-post systems common in Europe and America, and this has a significant influence on how people view their representatives. Under this proportional list system, citizens vote for a party, rather than for a specific person, and each party takes seats proportional to their share of the vote. The seats are filled from a list of preferred candidates selected by the party, and the party determines which politicians will represent which constituencies. There is therefore an absence of relationship between voter and representative because this list system means that people never actually vote their representatives in, and have no power to vote them out if they underperform. The only people with the power to appoint, transfer or dismiss an MP are those in the MP's party. As a result of this system, most people do not know who their MP is, and few would consider contacting them for assistance, primarily because representatives are comprehensively incentivised to please their party and have no obvious incentive to engage with their constituents. In South Africa there is also no expectation that MPs will vote differently from their party, so there isn't the motivation to influence MPs' voting habits as there is in other democracies. Moreover, as one research participant claimed: "Elections aren't battled on ideas in South Africa, they're battled on history".

Because of these issues, while presidents and prime ministers are generally universally recognised, the visibility of other members of parliaments/senates, including one's own constituency representative, can be extremely low. Digital platforms and tools were seen as an opportunity to increase the visibility of officials and reduce the space between citizen and government, enabling the two to interact digitally and empowering citizens to hold their politicians to account.

While digital tools and social media platforms have the potential to link citizens with their parliamentarians regardless of whether they know their representatives' names or obligations, there are many other important classes of information that can prove vital in enabling citizens and civil society to make themselves heard in legislative and policy activities. Knowing how politicians have voted, knowing what is said in plenary sessions and knowing what legislation is being drafted and scrutinised before a vote can provide opportunities to hold governments to account.

As seen in the previous thematic section, civil society groups concerned with parliamentary monitoring find the best channels they can to provide information. In Uganda this took the form of live-tweeting what was being said in plenary sessions in parliament. In Kenya, plenary sessions are published from Hansard via Mzalendo (as in the UK on TheyWorkForYou.com). These are the ways in which, the groups judge, they can provide the greatest value to citizens, who are thus able to follow the debates taking place in the official parliamentary chamber. In South Africa, plenary is described as "theatre" and "a few hours of hot air". The published edition of Hansard is barely used here, simply because no-one cares what "nonsense" is spoken during plenary sessions. The research participants in this study working in South Africa noted that Committee Sessions are where the real decisions are made, and information concerning these was thus far more important that any record of plenary discussion. Committees here are also the scene of public participation, where citizens and interest groups are able to provide oral and written evidence. PMG South Africa, the parliamentary monitoring group for the country, therefore focuses much of its efforts towards these committee sessions, employing monitors to attend meetings to take notes for publication online.

The differences in parliamentary structures and citizen understanding of those structures affects to what extent tech may be replicated across country lines. What is important for citizens of South Africa is less so for individuals in Kenya. The information needed by Nigerians differs from the information that Ugandans desire. The data collected for this study demonstrates that basic governmental and parliamentary/presidential structures, and citizens' understanding of those structures, will dictate how effective any replication of digital tools may be.

One of the most significant barriers research participants raised in building and implementing scalable digital tools for democratic engagement in Sub-Saharan Africa was the difference in parliamentary structures across different countries. This is unsurprising, given that transplantable parliamentary monitoring technology is notoriously difficult to effect even in developed countries with good open data. As noted above, every country has its own individual governance structures, and even where countries have similar parliamentary or presidential systems, there tend always to be nuances that mean code cannot easily be 'copied and pasted' from one place to the next without significant alteration.

These tools also rely on good, up-to-date data to provide solid foundations for digital tools designed to bring citizens and parliaments together, and on the whole, rely on parliaments themselves to produce this data. While again, this research detected a healthy appetite amongst civil society, policy-making, journalism and participation sectors for good quality information on the structures and behaviours of politicians within the legislature, it should act as a caution against assuming that good parliamentary monitoring digital tools can be replicated wholesale across borders with no prior research into the needs of citizens or structures of the legislature.

Access to Information and its quality

One of the most vitally important factors in the construction and maintenance of digital tools anywhere in the world is the quality and reliability of the data and information upon which such tools are built. Data is the oxygen of civic technology, without which digital tools are doomed to wither and die. Sub-Saharan Africa, however, is not yet considered an open-data-rich environment, and the bulk of the production and distribution of good data is carried out by civil society, a significant burden on an already under-resourced and often over-regulated section of society in the region. The World Web Foundation's Open Data Barometer places Kenya and South Africa at the top of the rankings of their Sub-Saharan neighbours, ranking 35th and 46th in the world respectively, but notes that neither country is yet demonstrating good, consistent open data practices. Nigeria ranks lower, at 70 in the world rankings, with Uganda trailing further behind at 90th, and it is interesting to note that these two countries have actually gone down in the ranking since the previous Barometer assessment, suggesting that either other countries have accelerated their open data efforts, or that Nigeria and Uganda have fallen backwards.

CountryStandardised Score
  1. Kenya (+7)

40.42

  1. South Africa (+1)

34.43

  1. Mauritius (-6)

26.3

  1. Ghana (+11)

25.96

  1. Tanzania (+2)

21.73

  1. Burkina Faso (+3)

21.63

  1. Nigeria (-3)

20.97

  1. Rwanda (-25)

19.62

  1. Ethiopia (-3)

16.14

  1. Uganda (-20)

11.59

  1. Malawi (-15)

10.51

  1. Namibia (-20)

9.08

  1. Senegal (-28)

8.74

  1. Sierra Leone (-14)

8.47

  1. Benin (-27)

7.32

  1. Botswana (-26)

5.89

  1. Cameroon (-29)

5.15

  1. Zambia (-22)

3.82

  1. Mozambique (-34)

3.46

  1. Mali (-23)

2.46

  1. Zimbabwe (-21)

1.87

Open Data Barometer 2016 for Sub Saharan African countries. Numbers are global ranking and change in global ranking.

The latter appears to be true of Nigeria, where public data that was previously accessible, in particular around health and education, has now been tied up with licensing and other regulations — or is simply no longer collected. This is surprising in a large economy actively courting e-commerce and digital development, and while some political and public datasets concerning finances and budgets have been made available, they represent only the tip of the iceberg at present. Country context is important in considering where resources devoted to creating good quality data are targeted, and in Nigeria, much of the focus revolves around the use of public finances.

Some digital tools have been developed in Nigeria in response to the common concern of pervasive corruption, notably BudgIT, which provides federal and state budgets, and Tracka, a portal through which citizens can track capital projects and report on potential issues. The organisation CODE engages citizens to 'follow the money' being spent by their politicians, and the Budeshi project run by PPDC monitors finance through open contracting, growing from its base in Nigeria to encompass public procurement in Kenya and Uganda.

These groups are making a heroic effort to produce good information and to make sense of it through reporting and creating infographics that enable non-specialist citizens to consume it. Much of the information is, however, painfully extracted from government rather than freely published, and while information available in any form is a step in the right direction, the persistent use of PDF format or paper responses to information requests reduces the usability of the information to others, as it cannot be easily copied or manipulated for other purposes.

The Nigerian government remains picky about which information it will freely publish, with the budget for the operation of the National Assembly withheld, and information on legislators and Assembly proceedings often patchy and out of date. The ShineYourEye website for parliamentary monitoring and engaging with representatives demonstrates how a site with good potential for informing citizens can suffer from a lack of reliable and quality information, as contact details in particular are frequently changed and cannot be updated without sourcing the new information directly from the politician in question. Additionally, the back-end digital infrastructure of the site was not suitable for easy administration by the group operating it, meaning that cumbersome procedures prevented new updates from being actioned quickly. These issues spell disaster for the usage and longevity of digital tools, as once individuals know a digital tool uses out of date information, they will be unlikely to use it a second time.

The Ugandan government demonstrates similarities to Nigeria in its reluctance to produce and publish good quality open data. There are several official sites professing to hold a wealth of open data, but in practice these sites contain very little that can be used to build a good picture of government activity, and have almost no data that can be reused to build other platforms. The government has passed legislation promoting open data and access to public information; however, current practice within government is completely opposed to this and access is tied up in lengthy and complex procedures and regulation that for the most part must be conducted offline.

This is not necessarily true across all departments, and there are individuals within the system that have championed greater public participation and consultation with regard to finance and planning. However, the research participants within this study emphasised the importance of the individual champions having not only the personal motivation, but the strategic position within government to achieve these reforms. The positive initiatives to involve citizens are not born of top down, structural commitments to better engagement; rather they have emerged as programmes that come from the deep personal commitments to change of the individuals leading them. The same personal touch was said to be the most effective way of sourcing information from government. The Freedom of Information Act in Uganda is often burdensome and complex to use, and therefore asking a personal contact within government is the only effective way of getting the information required. These are therefore admirable initiatives, but rely on individual interest and are vulnerable to changes in personnel. Change cannot be strategic in this model of governance, and citizens cannot be confident of their rights or expectations of government.

Empowering and effective digital tools for civic engagement are built upon strong data and information foundations. Without this very basic level of data infrastructure, it is difficult to envisage how digital tools can work effectively at low cost, as the only solution for the paucity of good quality data is to either produce it externally to government (within civil society), or to persuade government and parliament to produce it themselves. Solutions that rely on civil society production of data are vulnerable to funding cuts, personnel movement and political pressure, and divert civil society resource away from its primary role in holding government to account. As demonstrated in this section, governments are able to monopolise data, and as such, solutions to informing and engaging government and citizens together that do not rely on a steady stream of good quality open data are likely to be more effective than impressive platforms with functionality throttled by disruptions in data sourcing.

Public understanding

As discussed previously, different countries have different governance structures, and as such, citizens of those states have very different information needs. These needs, as this report has shown, are shaped by the institutional structures that govern societies, but are also driven by each public's general understanding of those structures. In order to engage in politics, public debate and governance, citizens must have a reasonable knowledge of these institutions. Without that basic level of knowledge, individuals will have no concept of the social contract between citizen and state, what the state's obligations are, or what they can reasonably expect the state to provide them with or consult with them on.

In Nigeria and Uganda, participants in this study noted that citizens have become almost bored with being angry about corruption and poor governance, and therefore often shy away from consuming media stories on such issues, or mobilising against them. In South Africa citizens know that appealing to one's constituency representative will likely amount to nothing because they have no incentive to please constituents. This apathy, a result of the structural environment, can mean that citizens don't engage even where they might be able to effect change. Moreover, some research participants believed that there are so many protests in South Africa because citizens do not know or have another way to communicate and engage with their representatives, or they feel that they haven't been listened to when they have tried to engage. They also expressed fear that protests in South Africa are progressively getting more violent with one CSO interviewee saying: "Some communities we've worked with have said 'If you burn a library down you know you'll get listened to'".

Unlike Nigeria, Uganda and South Africa, Kenyans are voracious consumers and sharers of political news, and the majority of citizens retain a belief in their own agency in public life, even in the face of corruption issues similar to those experienced in other countries within the continent. In Kenya, the underlying structural problems around corruption and cronyism have not dulled individual will to discuss politics or driven citizens to ignore it because they feel powerless. Kenyans are willing to be as outraged and animated about political stories that capture their imagination as ever.

However, there are interesting peaks and troughs in public interest which reflect the different levels at which issues occur — and the different types of political action with which the people respond. The recent devolution of many state functions to regional and county level has resulted in a huge rise in public interest in local politicians' activities, and there is some indication that at this level, people are punishing apparently corrupt or incompetent incumbents at the ballot box (high turnover of politicians at each election cycle is very common in Kenya, but even by Kenyan standards, the last county elections resulted in astonishingly high turnover).

This diversion of political interest to the local level has, according to the experiences of the research participants, somewhat insulated the national level legislature and executive from the wrath of the people, as citizens in Kenya are now generally very focused on tangible local issues, rather than the more abstract national policy and political behaviours. Scandals are therefore being punished at the local level, without corresponding monitoring and subsequent consequences being meted out at the national level.

In attempting to engage citizens in meaningful participation in governance and interaction with institutions, civil society groups and both traditional and new digital media organisations are on the frontline across the four case study countries. While each government stressed its policies on engagement and the opportunities it had created to enable citizens to participate, in reality many of these opportunities were either tokenistic or deliberately difficult to attend. The third sector and the media bear the brunt of educating citizens about their rights and informing them of the activities being conducted in parliament and government. The participants of this study stressed that in doing this, an engaging narrative is essential in capturing the interest of citizens. Political and official information in itself is of little use to citizens, who are much more engaged by stories that illustrate how the issue at hand can or does affect them on a personal level. One participant noted:

Of course, it's true anywhere that pieces on politics and government need to be centred around a good story, but in my experience, nowhere is it more true than here in Africa. People here need a compelling narrative, and need to know how it affects their lives.

This is something that was repeated by many of the participants, and demonstrates that public understanding of government, and engagement with it, is interwoven with the way that information is presented to them.

There is an educated and cosmopolitan elite who are interested in governance in each of the case study countries; however, the majority of citizens are rural and have a lower level of education than those who are prominent on Twitter and who are able to speak publicly and articulately about political life. It is the more silent, less educated and less affluent majority that provide the majority of votes when it comes to election time, and who therefore have the greatest influence over who takes power in the national government. Political information thus needs to be simplified, contextualised and presented in a consumable format. In the case study countries, the most effective NGOs and media operations working on democracy and openness in government are the ones that provide interesting and easily consumable political information, and that manage to marry the analogue and digital outreach they perform. While digital may be the future, the present is still dominated by the radio and town hall meetings. As one research participant from a South African public interest news organisation said: "We produce posters and that's one of the most successful things we do. Low tech can work just as well as high tech".

This may be a more significant finding than it initially appears, especially when considering how governments themselves in the case study countries present the information that they generate. Parliamentary information on the official websites is sanitised, often patchy, and written in impenetrable language for the majority of citizens that have no legal or parliamentary training. Social media posts are often short, in functional language, or tell no real story behind the activity being shown. Government statements are often dry and rambling. Outreach by government itself to rural areas is sporadic and notoriously difficult to participate in. It is only through the civil society and media storytelling intermediary (or infomediary) that citizens are able to understand what is really happening in government.

Digital parliamentary monitoring tools

In each of the case study countries, some form of parliamentary monitoring tool was in operation, run by a civil society group with the aim of providing high quality information on the members and activities of each national parliament. The aims of these sites are to raise awareness of parliamentary and policy activity amongst citizens, the media and civil society, who should then be better informed and well placed to campaign on certain issues; and to hold the government to account for its actions and decisions. The core belief that drives these sites is that sunlight is the best disinfectant, and that in order to tackle corruption and cronyism, the electorate must first be able to see clearly the internal machinations of parliaments and public institutions, and further, be able to use that information to effect change. The organisations who run these parliamentary monitoring tools pride themselves on being non-partisan and neutral, and this study found no evidence that they were viewed otherwise — a finding that was not a foregone conclusion.

It has been shown in a previous report by mySociety (2015) that individuals who use digital parliamentary monitoring tools find them useful, and that the users of these tools genuinely believe that politicians would behave differently if the information contained on the sites was not readily available to the public. At the very least, this demonstrates a positive, if somewhat passive attitude to the beneficial effects of parliamentary monitoring tools and their ability to improve governance.

The visibility and effectiveness of the platforms in operation in each case study country were variable. Kenya's platform Mzalendo, and South Africa's twin platforms PMG and People's Assembly were significantly more widely known and used than ShineYourEye in Nigeria, and ParliamentWatch's website in Uganda was not as widely used as the organisation's corresponding Twitter feed; however, the lower usage of these latter two platforms can be explained by the data environments discussed above. At the fundamental level, the majority of research participants in this study used the relevant sites in their country, and relied upon them to do their own jobs. The PMG and People's Assembly websites in South Africa were confirmed by the participants of this study to be vital to the majority of organisations working in politics and policy.

In South Africa, significant traffic and subscriptions to the organisation's premium website content come from within parliament, and in Uganda parliament staff were the second highest user group of Parliament Watch, behind NGOs. In fact, when the South African parliament's subscription lapsed and was not immediately renewed, parliamentary researchers clubbed together and paid for the subscription out of their own pockets, such is the necessity of the PMG website to their work. In Uganda, Parliament doesn't have an internal communications system where parliamentary staff can access minutes/information from committees they were unable to attend, so they use ParliamentWatch to fill this gap. In Kenya, South Africa and Uganda, journalists and NGOs would simply be unable to do their jobs properly without the digital parliamentary tools provided by civil society.

A primary issue with the development and usage of parliamentary monitoring sites (and their corresponding social media) is that they are overwhelmingly visited and regularly used by the educated, affluent and cosmopolitan elites in each country, many of whom are using the sites in the course of their own professional activities. This can clearly lead, very logically and reasonably, to a concern about the ethics of building and supporting digital tools that are being used by an already engaged and affluent minority. However, this line of argument must be followed through to a conclusion that rejects the assumption that these tools only benefit the elite. The fact that so many users of these tools rely on them for their own professional roles, the majority of which are in legal, civil society, campaigning, environmental, media and academic spheres of activity, means that these parliamentary monitoring sites have a trickle-down effect that is impossible to calculate.

Discussing how the information on these sites is used, participants cited news stories, radio call-in shows, legal challenges on behalf of communities, and the submission of vital evidence to change policy as just a few ways in which the impact of these sites is amplified beyond that which can be calculated from usage statistics only. One radio DJ interviewed discussed how, at least once a week, while on the air, he would monitor what was being said in parliament (being live-tweeted by ParliamentWatch.ug), and would use the topics being discussed to invite the public to call in and share their views. Many people would have no idea what was going on in parliament that day without this show, and fewer would have the chance to discuss it in any meaningful way. While the Twitter feed for this parliamentary monitoring site had 54,000 followers, the timeslot on the radio station in question had an average of seven million listeners.

Lower usage of parliamentary monitoring tools can suggest that they are not effective. However, the impacts can be further reaching than it is possible to determine. The trickle-down effect discussed above is a significant consideration in understanding how important such digital tools can be. Another is the potential for cultural change. The group that runs ShineYourEye in Nigeria admits that there is not a current culture of parliamentary monitoring in the country, and that this is something that they are actively attempting to change with the platform. A chicken and egg argument could ensue, where it is not clear whether the platform should come first, or the culture change; however, the platform is already being used by a minority in positions where trickle-down benefit can occur, and so is logically of societal benefit even if citizens themselves are not using the site for their own individual needs at present.

One of the key success factors in running an effective parliamentary monitoring tool is a strong relationship with the parliament and its staff. The groups running Mzalendo, ParliamentWatch and PMG/People's Assembly have spent years forging mutually beneficial relationships with key politicians and administrators within parliaments, to ensure a good flow and quality of data, to enable access for them to conduct their monitoring activities, and to promote the benefits of parliamentary monitoring to the institutions themselves, as well as for citizens.

These relationships require a constant balancing act. Parliamentary monitoring groups must avoid appearing confrontational or partisan, and must avoid preaching too much at any one time. As one research participant put it, "Once you name and shame [politicians] it's impossible to engage, the relationship is broken". Organisations must also press the benefits of their work and subtly promote the need for good data, without making unpalatable demands or seemingly embarrassing the politicians or their civil service. Through incremental, positively framed interactions, these groups have opened up parliaments to the public in a way that no other organisations have been able to achieve.

This study found that digital parliamentary monitoring is a positive and beneficial service in the case study countries, and has much deeper impact than can be reasonably measured from looking at website statistics in isolation. Parliamentary monitoring platforms are well used by those in policymaking, media and civil society, who are able to translate the often oblique language of parliament into understandable and consumable information that citizens are interested in. These sites provide a unique service unavailable through official websites or other channels, and should be seen as a key digital tool for promoting better governance and democratic engagement, that acts as a vital cog in a wider infrastructure of information exchange between citizen and state.

Social media

No analysis of digital interactions between citizens and public institutions in 2018 would be complete without a discussion of the use of social media. In little over a decade since its inception, social media has become one of the primary channels of communication for individuals all over the world. While initially designed to enable individuals to interact with a pre-existing offline friendship group, social media now provides for interactions across all walks of life, and with anyone, anywhere, irrespective of distance. All that is required is an account and the means to use it.

Social media usage in Sub-Saharan Africa mirrors the figures on smartphone and hybrid phone usage, so if someone has the ability to be on social media, the likelihood is that they are. In 2017, 49% of South Africans use WhatsApp and 46% use Facebook. In Nigeria the equivalent number is 41% for each service (Statistica 2017a and 2017b). In Kenya, Whatsapp is used by 12 million (24%) and Facebook 7.1 million (14%) (iFreeKe, 2017). While there is no exact equivalent statistic for Uganda, a rough estimate places the number of social media users around 3.8 million (9%) (Daniel, 2018).

WhatsAppFacebook

South Africa

49%

46%

Nigeria

41%

41%

Kenya

24%

14%

Uganda

9% (general social media use)

NA

Participants in this study with a personal or professional interest in government and parliament all noted that WhatsApp is now the leading platform for sharing content, above Facebook which is seen as more static, and above Twitter which is used less prevalently outside of urban areas. WhatsApp usage in Kenya is significantly higher than within the other case study countries, with individuals often in a very high number of different 'group chats', commonly with people that they have never met before. While radio is still king across the whole country in terms of disseminating political news (with newspapers coming in second), anyone with access to data is most likely a member of 100+ WhatsApp groups, some of which will be completely devoted to politics (whether that be at a local or national level), and sometimes broken down into thematic or geographic groupings.

Some WhatsApp groups even form webs of related political groupings devoted to strategising how to achieve certain goals such as accountability at the local level or strategically voting for a new candidate, across larger geographical areas and multiple loosely linked groups. Elected representatives themselves are often in these groups, and citizens post questions and ask representatives to comment on issues, as a way to keep them accountable.

There exist both informal organic groups set up by private citizens, as well as groups implementing well thought-out usage strategies that have been instigated by NGOs and journalists. These groups often reach maximum capacity group sizes on WhatsApp (as many of the interviewees told us, WhatsApp allows for a maximum of 256 members per group chat), and therefore overspill is now being managed by many group administrators using Telegram. This maximum capacity also leads group administrators to post messages from one group to multiple others, creating a chain effect.

This barrier to wider inclusion may eventually mean that Kenyans migrate from WhatsApp to another platform that better suits their needs in the future, but currently, there is no indication of mass movement to other platforms. WhatsApp has replaced traditional text messaging in Uganda and Nigeria among data-users, and is used both as a sharing platform and as a direct form of contact with individuals. Its usage in South Africa is prevalent, but tends towards group chats in which all of the members are known to each other, whether this is professionally or personal contacts. Indeed, in South Africa many policy professionals belong to a group that is used solely to disseminate press releases from parliament, as this is the fastest and most reliable method of getting this information.

In light of these findings, it is clear that WhatsApp is one of the most important factors in thinking about how people are communicating official information and political gossip, as this is the platform where people are sharing news, gossip, opinion and content. This means that good, legitimate information must be shareable in WhatsApp-friendly formats in order to have any chance of impactful penetration. Many of the more educated political enthusiasts will get information or gossip from WhatsApp and then go directly to a more reputable source to fact-check that information. A lot of these people in the case study countries use Mzalendo, PMG and ParliamentWatch in this way, to check things they have seen elsewhere before sharing them further or repackaging them for wider consumption. An average citizen with no political, policy or governmental profession would be less likely to do this, in particular within a group comprised of their trusted local community. Therefore, it seems logical that high quality content should be easily delivered to them in the shareable WhatsApp format that people currently have a digital preference for.

In consuming and sharing information on government and politics, Twitter tends to be used more by the urban or cosmopolitan elites in the cases examined, and often by the diaspora, especially the larger diasporas of Kenya and Nigeria. These individuals tend to have higher levels of education, experience of living, studying or working internationally, and are comfortable being vocal and confrontational on the platform, in particular in the Kenyan case. As diaspora, many of whom are living in Western democracies, they remain eligible to vote, and remain actively engaged with their domestic social networks. Facebook usage, while high, tends to be considered less fashionable in terms of active discussion on politics and governance, and while citizens in more rural areas still consider Facebook to be 'the internet', citizens in urban areas have migrated away from it, preferring Twitter and WhatsApp for politics, and Instagram for personal use.

The use of social media by citizens to interact with public bodies and public officials is not yet as prevalent in Sub-Saharan Africa as it is in the UK or USA, but it is travelling in the same direction. Many politicians and government departments across the four case study countries now have Facebook and Twitter profiles; however, the maintenance and monitoring of these accounts, and the interaction through those channels, varies widely. Those politicians that do have social media profiles and who are actively responsive to messages from the public tend to be younger members of parliament more familiar with the technology and more comfortable with an element of transparency in their work. One South African research participant summed up the importance of social media for politicians: "The political party that finds out how to engage 18-25 year olds (those who currently vote the least during elections) on social media will undoubtedly win the next general election in South Africa".

In Uganda, ParliamentWatch's strong working relationship with parliament enables them to provide training for MPs to get online and interact with citizens through social media. Online accessibility to representatives is in no way complete in Uganda at this point, and many of the older members of parliament refuse to adopt digital ways of working; however, the young population in Uganda is vocal in demanding that MPs and official departments are accessible in this way, and this has helped politicians to buy into the kind of training that ParliamentWatch provides. Mindful of how citizens use social media, this organisation also creates a lot of shareable video material that can be distributed easily on WhatsApp, and they repackage parliamentary information into more consumable blog and visual formats. The shareable short videos on political topics are highly valued by groups working with communities that have low literacy levels, as this format is more accessible to those who perhaps only speak English rather than read it, and the files are small enough to be shared widely via social media apps. One Ugandan journalist from a mainstream newspaper told us that in the absence of Ugandans being allowed to protest:

"Social media is the one thing that is working really well. It's the one disruptive force that seems to be getting tangible results, where people are able to share their frustrations and send these directly to the people in charge in front of the whole world."

Many of the Ugandan research participants we spoke to for this study shared the belief that asking for information or complaining on public social media channels and websites is having a positive effect on response rates from government departments and politicians, and have had personal experiences of this working well. However, these experiences are still few and far between and are likely to be enjoyed by citizens who have more prominent social media profiles and who are prepared to put their head above the parapet (something Ugandans in general are still afraid to do in fear of reprisals such as prison sentences, destroyed homes and even death — freedom of expression continues to be an issue in Uganda). In addition, the Ugandan government has shown hostility to social media, previously shutting down social media provision on election day in 2016, and in 2018, introducing a new 'social media tax' that will make it more expensive for citizens to share political information through these channels.

In Kenya, politicians are very aware of Twitter, and while most are not very responsive on the platform, research participants in this study explained that the politicians find it a useful tool to gauge the attitudes of the populace. Twitter seems to be a safe space for people to support opposition views, and seems the primary place for any critical dialogue inside or outside of parliament. This is because the current official opposition party in Kenya is no longer trusted to oppose anything, following some controversial political moments earlier in 2018 and the subsequent failure of the newspapers to examine these and related stories critically. Indeed, one participant in this study went so far as to state that "in Kenya, Twitter isthe opposition party".

People are very willing to complain directly to MPs on the Twitter platform, and there is clear evidence that so called 'Twitter storms' have been responsible for the government rolling back on policy or making public comment. A journalist interviewed for this research described a recent case in which one of his colleagues was allegedly detained by government actors. The ensuing Twitter fallout was considered instrumental in prompting senior officials to make a statement on this allegedly illegal detention on Twitter, and release the journalist very swiftly.

The users of Twitter in Kenya identify themselves with the hashtag #KoT (Kenyans of Twitter), and influential individuals are able to mobilise significant support for issues via the platform. As discussed above, Twitter is predominantly an elite tool, but it is this very elite usage that has caused politicians to become increasingly responsive to criticism or campaigning conducted on the Twitter platform. With an invested, articulate and efficacious user base, Twitter users in Kenya represent a threat to government, and government responds accordingly. An interesting factor in these governance-related social media movements is that a lot of politics and policy enthusiasts are using the parliamentary monitoring site Mzalendo to try to leverage even better responsiveness. These individuals have taken to tagging Mzalendo in their complaint tweets, as they believe this will make politicians more likely to respond in some way. Anecdotal evidence collected by this study suggests that this is, to some extent, true.

Politicians and institutions in South Africa are similarly variable in their adoption of social media. While some public individuals and departments have embraced social media, there is no uniform approach taken by government or parliament to communicate information or be responsive to citizens via social media. Some institutions prefer to avoid social media altogether, and the Democratic Alliance party (the second largest party in parliament) currently prohibits politicians from tweeting unless they have cleared the tweet with the central party office, a cumbersome policy that in effect prevents the use of Twitter by its elected members.

The use of social media in Sub-Saharan Africa is rising, and this study has shown that populations are eager to engage with governance structures and politicians through these new formats. There has, however, been a reticence on the side of the public institutions to interact with citizens through this particular medium. Often, older elected individuals or officials simply do not want to learn how to use the new technology, and the majority of public institutions in Africa lack internal expertise in the use of social media for citizen engagement.

Some parties and administrations have called social media a distraction, and others have deliberately refrained from using social media as an official channel in order to avoid difficult or critical dialogue. In Uganda, the purpose of the new tax on social media usage has popularly been claimed to curtail citizen engagement with governance. While the more confrontational platforms such as Twitter may be under-utilised by parliamentarians and public institutions, the acquisition, repackaging and disseminating of public information through digital means and by civil society and new media organisations is growing all the time. Citizens across the case study countries are increasingly finding their political and governance information online, and the significant youth populations are becoming louder in their demands for a new way of interacting with their officials.

The enduring significance of social media in citizen and public institutional interaction means that the need for information to be packaged in a shareable and consumable format is clearer than ever. Simply having an impressive website will not ensure citizens have good quality channels for acquiring or exchanging information. Using only Twitter to promote information is likely to disenfranchise more than half of a country. Used in isolation, these social media channels are somewhat patchy, but they are now being used in a combined effort that blends organic fact-checking with parliamentary monitoring and other trusted news sites, to make more citizens more informed than ever before. The challenge for organisations wishing to support these democratic efforts lies in building tools that are relevant to the evolving digital behaviours of citizens in the region.

Traditional media

A discussion on traditional forms of media might be considered out of place in a report concerning digital means of interaction between citizen and state; however, offline forms of information consumption remain the most popular in Sub-Saharan Africa as a whole. While it is likely that digital consumption of information in the region will rise to near saturation point over the next ten years, currently, the majority of citizens develop their understanding of government from newspapers and the radio.

Conscious of the growing migration to digital means of information consumption, traditional media houses are also developing their own digital presences, the better to reach users in both formats. The main newspapers in the case study countries all have online presences, and are mostly mobile-optimised, since the majority of internet users access content through a mobile device. Journalists and broadcasters for these traditional media houses are capable of generating huge social media followings online, and are conscious of the need to share stories digitally, in particular to reach the large youth populations that have rejected analogue forms of information consumption. Several journalists were interviewed for this study, and all were conscious of using digital channels to distribute their stories as widely as possible, but also as a means of securing information first. Digital has enabled the early adopters in traditional media to find and break stories ahead of the pack, and keeping digital channels open to receive information is now a very important part of their role.

There is, at this time, a clear body of professionals in traditional media working with a foot in both analogue and digital worlds. While the end product may be tangible, the methods of producing it are no longer the same. The information that journalists work with, from or about institutions and officials, is increasingly received digitally through social media, through private email, through digital streaming or through group messaging. Journalists are now often sourcing information digitally and distributing it via the radio or a newspaper. Earlier in this report, the marriage of analogue and digital was encapsulated in the practices of a radio DJ in Uganda who used Twitter to follow live parliamentary debate, and then used that information to engage his average of seven million listeners.

In addition to websites of the long-standing traditional media, new online-only news and politics sites have sprung up in the case study countries in the last five years. One of the most impressive is based in Kenya. Judging by the outpouring of respect and affection by the participants of this research, the only universally trusted source of traditional news and opinion amongst policy, democracy and governance actors in Kenya is considered to be 'The Elephant', an online news site established in 2016 to 'speak truth to power' and to fill the gap in non-partisan reporting and investigative journalism in Kenya. The quality and reliability of the site was mentioned in every interview conducted in Kenya; it has a broad and expert team led by John Githongo, possibly the most famous anti-corruption activist and government whistle-blower in Kenya. The Elephant have recognised the need to create different types of digital content and have a recording studio where they produce podcasts, vlogs, and two-minute video 'factoids' to share on social media platforms; they also regularly run Twitter chats that trend nationally.

Contributors to this research have claimed that the parliament in Uganda is practised at trying to suppress information it does not want in the public domain; however, the rise of social media and online parliamentary monitoring tools has started to disrupt its absolute power to prevent instances of corruption from emerging, and citizens are actively engaging with the information coming out, as well as demanding its publication. While it was reasonably easy to prevent physical papers from being printed, or to remove the physical ability of a radio station to broadcast, it is more difficult to prevent the dissemination of information online. An example given by a participant in this study detailed a recent incident in which information distributed about parliamentary practice enabled citizens to remove their area's incumbent representative from office. In the course of renewing the constitution, the President asked parliament to remove the limit on his term. The parliament allegedly put pressure on the media to refrain from reporting this; however, media houses proceeded to do so, with the content being shared widely online and through social media, and as a result many citizens became very vocal about their opposition to this change to the constitution. MPs were allegedly bribed to vote in favour of the change, again with pressure exerted on the press to avoid reports appearing about the vote. The press rebelled again, and citizens again expressed their discontent. The consequences of this were played out in a recent by-election, in which the incumbent was voted out specifically because he chose to take the bribe and vote against the wishes of his constituents. This is a rare victory for people power versus the government in Uganda.

Across this study participants claimed that mainstream newspapers often sensationalise and misreport political information in order to sell papers. Organisations involved with running digital parliamentary monitoring tools felt they were in a privileged position, because they don't have the pressure to sell papers to keep afloat: they can afford to provide accurate political information. The lack of easily accessible information from government also fuels inaccurate news stories in the media - for example a headline news story about Kenyan MPs' pensions rising by 700% was factually inaccurate due to the lack of disclosure from parliament, not because of bad journalism. As one research participant put it: "When the media gets it wrong, the whole country gets it wrong" Examples like these reiterate that 'fake news' isn't a recent online social media phenomenon; it has been an issue for offline media for a long time too.

The importance of traditional media cannot yet be ignored, and it will continue to be influential for many years to come. However, the lines between traditional and new media are blurring, and whilst many people will consume only analogue information, that information is increasingly being generated and fact-checked digitally. This means that high quality traditional media is becoming more and more dependent upon high quality digital media, and ever more dependent upon civil society groups, parliaments and institutions producing good quality data and information with which to build those digital media platforms.

Political will

Across this study, a fundamental need was found for political and institutional engagement in providing timely, good quality and relevant information or responses to citizens, in order for the feedback loop to be closed and for citizens to feel that their concerns and needs were being heard by the governing administration. As has been touched upon in many sections in this report, government must 'play ball' in order for infomediaries to package and disseminate information and for digital to be a two-way channel of interaction for citizens to hold their governments to account.

One of the key issues preventing institutions from providing better information is the desire to avoid scandal. Corruption remains embedded into many public functions in Uganda and Nigeria, and has not been completely stamped out in Kenya or South Africa, even though further checks and balances have been implemented. In each case study country, there are significant issues around procurement, contracting and budgetary transparency, in particular concerning the large-scale infrastructure development that is happening across the continent at the moment.

Significant sums of money are pouring into Sub-Saharan Africa from a number of international sources (China, Europe and USA), and each of these sources of funding has different requirements attached. Significant taxpayer money is also being contributed to these large structural projects, but there remains an opaqueness to the award of contracts and the payment of funds. These public procurement and contracting exercises have historically been extremely vulnerable to corrupt practices by parliamentarians and government officials; however, it is difficult for civil society and media groups to gather the necessary information on parliamentary and departmental involvement in the processes to trace where the money goes. Parliaments are understandably reluctant to provide information that demonstrates wrongdoing on their own part.

There also remain cases where politicians simply do not understand why 'their' information should be made public, especially on sites hosted externally. This attitude isn't necessarily drawn from the desire to avoid scandal, but more from a lack of understanding of why citizens would want and care about this information, and the 'us and them' mentality of some politicians. One of the Nigerian organisations included in this study has appeased those with this attitude by creating transparency portals on existing government websites, as hosting the information on an external site was seen by the politicians as a step too far.

As such, it is difficult to envisage a point at which parliaments will genuinely discover within themselves the political will to provide complete and open information on their activities, unless attitudes to corruption and transparency within government shift dramatically. This is not to say that this is not likely; Kenya, South Africa and Nigeria have all demonstrated movement towards better governance structures and greater information provision. This movement towards more transparency simply requires more time than Western donors would perhaps like to contemplate. Generational change will no doubt push the needle on this, as will it instigate further migration to digital working by default.

Another facet of the need for political will lies in the different expectations of where responsibility for the production and distribution of information for citizens lies. Civil society groups have been historically burdened with conducting much good governance work in Sub-Saharan Africa over the last 70 years, and as such governments have become accustomed to this arrangement. As progress into a more digital age goes forward, and as governments move towards greater transparency and interaction with citizens, those functions can no longer reside completely outside of public institutions.

A good example of this is in South Africa: although the parliament is obliged to make committee proceedings a matter of public record, in reality this function is performed privately by PMG, who organise and finance the minute-taking of committee meetings, and digitise and distribute the records. As such, there is little inclination to take on this work in-house, even though this should logically form part of the administration's duties. Bringing this function in-house and publishing the records digitally would enable civil society organisations such as PMG to focus more work and resources on the actual content of the parliamentary record, rather than simply the production of it; however, the excellent job performed by PMG has seemingly fostered complacency in the parliament and its administration.

This research consistently found that political will and support is key in creating a digital environment in which citizens can consume good quality information about their government and respond accordingly. Where there is blockage due to corruption or a lack of expertise, the fate of digital platforms to facilitate citizen and institution interaction looks to be gloomy, as without one or more champions inside institutions driving the need for change, it is easier for administrations to continue as they have until this point.

Even social media as a channel for citizen and institutional interaction is vulnerable to a lack of political will, as individual politicians, their parties or their administration create rules for its usage or avoid it completely in order to avoid difficult questions or criticism. Where individuals within institutions have bought into digital as a means of communication with citizens, positive outcomes have been identified, as detailed by sources in the Ugandan administration who have experienced first hand the improvement in citizen engagement and openness within a single department governed by an open government advocate. However, more of these individual champions are needed in order to nudge creaking institutions into the open governance models of the 21st century. Only where there is political will to be open and engaged, can digital platforms to enhance engagement flourish.

Critical success factors

Drawing together the thematic discussion above, it is clear that there are certain indicators that are likely to affect the success or otherwise of a digital engagement tool in Sub-Saharan Africa.

This study has shown that where any of these factors are weak, digital democracy tools are likely to be ineffective at best, and possibly harmful at worst, in helping to perpetuate the idea amongst citizens that engagement does not work.

Recommendations

The purpose of this research was to better understand how digital democracy tools and wider online information and engagement channels are operating in Sub-Saharan Africa, and how information is being packaged, disseminated and consumed. The hope is that understanding these processes should enable organisations like mySociety to build better and more effective civic technologies, and should inform new and future digital democracy projects on the opportunities and risks associated with building and implementing these solutions.

This research project has been conducted with several key audiences in mind, loosely grouped as civic technology organisations, development funding and philanthropic bodies, policymakers, and grassroots democracy organisations. As such, the following recommendations may apply to one or more of these groups.

1 - Conduct an in-depth scoping exercise in-country before committing to fund, build or implement a specific solution.

The growing importance of preliminary user research signals a turning point in how digital tools are now developed; however, the lack of structured and consistent research at the discovery phase of product creation remains one of the biggest factors in creating ineffective digital tools. A short field visit by technologists to talk to relevant groups and stakeholders would provide necessary information to ensure that tools are built to suit the landscape. Funding bodies should require groups to perform these early exercises, and assume that plans are likely to change from the initial project idea. Civic technologists should also be empowered at this stage to reject the idea of building new technology if the local circumstances suggest that a digital tool, or the tool agreed with funders, may not be the best solution.

2 - Work with partners in-country that have a good working relationship with parliament.

The most effective digital participation tools examined in this study were those that were not built in isolation, but that were very much on the government's radar and formed part of a broader strategy of engagement between the civil society group running the tool and the government it was designed to monitor or facilitate communication with. Where groups running digital platforms demonstrated a good working relationship with parliament and institutions, access to the necessary data to make the tools high quality and up-to-date became easier, and further opportunities to open up government and governance to citizens appeared to follow. In addition, partners in each country must have a certain amount of expertise both in policymaking and in the development and maintenance of digital tools, as well as a realistic understanding of the level of work and funding required on an ongoing basis to keep the tools running. Where groups with this profile do not exist in countries identified for digital development work, interested parties should focus initially on developing the capacity of groups to take on this role, investing in skills and sustainability, and supporting smaller interventions and engagements with parliaments in order to build a foundation of trust and good information, before implementing digital tools.

3 - Make peace with solutions that aren't necessarily scaleable or replicable.

One of the great hopes for digital democracy tools and open government processes was that tools would be scalable, so that the wheel does not end up being reinvented in each individual country. Indeed, open data and open government was called for by Barack Obama on his very first day in office. The vision was that open source technology could be easily transplanted from one country to another using template tools and local open data produced by government. When this idea is applied to parliaments, however, it does not necessarily stand up. The nuances of governance structures in each individual country reduce the replicability of digital tools to the extent that many require significant re-coding in order to accommodate what can often seem to be superficial differences. The parliamentary monitoring tools examined for this study all contained significant differences in structure and usage, as a result both of the structural environment, and the behavioural traits of citizens in engaging with it. While the basic code for ShineYourEye in Nigeria, Mzalendo in Kenya and People's Assembly in South Africa is similar at its core, the three platforms are very different. ParliamentWatch Uganda is different still, and the South African sister site to People's Assembly, PMG, responds even further to the unique needs of citizens and civil society.

This study found a common need for the fundamental data across the case study countries, but very different needs in terms of how that data is presented, packaged and disseminated, meaning that one design of digital tool cannot realistically be expected to be universally useful. While the desire to avoid recreating tech that already exists is a sensible one, this study suggests that it may be a false economy, as the primary common ground is the basic data that feeds these digital tools, with institutional structures differing in each territory. Spending the necessary funds to get a unique or hybrid digital tool right is likely to have far greater beneficial impact for citizens, and be easier for civil society groups to administer.

4 - Focus on arranging a reliable data source before trying to build.

As touched on in recommendation 3, reliable and good quality data is the foundation of high quality digital democracy tools, especially those with the function of providing information and communication channels to citizens. While building new tools can at first seem like the most desirable path to take, producing more tangible outputs, in reality, this course of action should only be taken if the data environment is solid. If the data required for an effective tool is not currently open as standard, or of high enough quality, then investment in human and financial resources to bring this level up is the best possible alternative action. This should be a serious consideration for any organisation intending to fund the development of digital tools for democracy, and should not be considered an unattractive option simply because it may require a longer term outlook on achieving impact. Trying to build digital tools for engaging with governance on poor data is akin to trying to build a house on sand, and in the absence of strong foundations, those tools will quickly become obsolete and add to the collection of digital tools that have fallen by the wayside because of poor planning. The global civic technology community has in most cases been vocal and honest in circumstances in which prescribed digital tools are unlikely to be effective; however, private and for-profit digital organisations simply abiding by a contract are less likely to consider the longer term suitability and sustainability of such builds. Therefore it is likely to be more beneficial to large scale digital development projects to work with civic technology organisations already familiar with the wider social and political landscape, rather than with private contractors supplying digital services remotely. Where good data is not yet available, the highest priority for investment should be in creating the conditions for that data to be produced, whether this is through working with governments and parties to agree to conduct this work, or working with (and funding) civil society groups to circumvent institutions that are unwilling to produce data.

5 - Ensure ongoing, stable funding for maintenance and growth.

In Sub-Saharan Africa, as in the majority of other countries in the world where civic tech has been implemented, useful and impactful digital democracy tools simply cannot endure beyond an initial start-up grant without further funding being secured. While $100,000 will fund the necessary expertise to build and implement a useful digital tool over a limited time period, if the initial grant expires without sufficient ongoing financial support, the likelihood is that the site will quickly become out of date and obsolete. Maintenance funding is necessary for the groups operating the sites to solidify their work with parliament and to grow the usage of the tool, and is also necessary for the groups providing the digital development expertise, if these groups are not one and the same. Successful sites require constant attention to fix bugs and respond to changes in data sources at the back-end of the sites; they need developers with the ability to amend platform features quickly and easily to respond to parliamentary activity and user needs; and they need staff or volunteers willing to conduct research into the platform's effectiveness. Several respondents to this research cited the short-term nature of funding for digital tools as a reason why some have fallen out of popular use. When digital tools for democracy become obsolete, it means that these tools are often forgotten and then re-invented, that expertise is lost, and that citizens lose faith in the ability of digital tools to meaningfully enable them to engage. The unfortunate fact is that the majority of civic tech and open democracy NGOs will never be fully self-sustaining, and these groups will inevitably close projects and digital tools that no longer attract funding, in order to preserve their existence through new projects that will attract funding.

6 - Ensure digital tools are able to integrate as much as possible with social media.

In the process of building contemporary digital tools to enable greater scrutiny of public institutions, significant attention needs to be given to how easily the information can be extracted and shared via social media channels. Citizens across the case study countries were vocal about the need for information accessibility within those favoured channels, without the need to leave the social media environment to load webpages or access different channels. Information on parliamentary activity that was repackaged for easy sharing and viewing within WhatsApp, for instance, had greater potential for citizens to see it than information simply tweeted or shared as a link to a webpage. While tools or websites do not need to exist on social media platforms, or directly feed into them, there should at the very least be the capacity within the operating organisation to curate and repackage content to post to these channels. Shareability of information is one of the key requirements of effective digital democracy tools in Sub-Saharan Africa at present.

Further research

This report has demonstrated that the ways in which citizens are acquiring, consuming and sharing digital information on governance and politics are varied and very dependent upon local context in Nigeria, Uganda, South Africa and Kenya. It has demonstrated that information created by parliaments is often too opaque for general public consumption, but that it is of significant interest once repackaged and disseminated by civil society and media groups.

The migration to consuming political content digitally is continuing apace, and citizens want to be able to share via their existing digital ecosystem. And it has been shown that these information flows are only possible in environments with the right data and access. This report is not an exhaustive account of digital governance and participation, and as stated in the earlier sections, it examined these issues through a parliamentary lens with a focus upon politicians and parliamentary information.

While this report illuminates several important overarching factors in this process, it has only scratched the surface with regard to the detail of other factors, such as individual citizens' lived experience, the legislative environment shaping information sharing, civil service digital interactions with citizens, and the monetary incentives for and against opening up governance. There are therefore limits to the ultimate conclusions that can be drawn, and the following points outline areas of research that would strengthen the knowledge available to assist in driving forward digital governance and citizen participation in the region:

  1. Based on the evidence collected, it is clear that citizens of all of the case study countries generally had an active interest in official, political and parliamentary information. However, the limited scope of this study prevented further inquiry with citizens concerning their individual behaviours, requirements and emotional experience with digital interaction with governance mechanisms. In particular, it would be useful to understand better the ways in which citizens and parliamentarians are currently conducting individual dialogue digitally through platforms such as WriteInPublic or via email or direct messaging. While the number and breadth of interviews conducted for this study was sufficient to illustrate the broad landscape of online interaction between citizen and state, the findings would be significantly enhanced through large scale interviews and focus groups with a wide range of ordinary citizens in each country.
  2. The legal environment in each case study country necessarily influenced what official information concerning the activities of parliaments and governments could be produced and published. On more than one occasion during the study, legal instruments such as an 'Official Secrets Act' or similar were cited as a reason why certain information was not available to the public. Additionally, the robustness and power of legal instruments with which to acquire information was also found to be variable and vulnerable to abuse. While each country had legal channels through which to request official information, in practice these were rarely fruitful. Further research into the legislative frameworks governing data and public information in Sub-Saharan Africa would provide valuable intelligence when embarking upon any data- or information-based service for citizens, and identify where investment could be targeted to push for greater reform in information rights.
  3. While this study primarily focused on the digital production and dissemination of parliamentary information and the interactions between parliaments and citizens, the interaction and information consumption between citizens and the layers of local and national government is an equally important sphere in improving and opening governance. In several interviews in this study participants noted how difficult it is to monitor the progress of works agreed in parliament, but delivered by the civil service, and it has been shown in many previous studies that this is the point at which the most pernicious corruption tends to occur. While there are a number of organisations conducting excellent monitoring work in this area, there are few studies using a specifically digital lens to understand how and where information can be sourced, and how it can be repackaged for consumption by citizens.
  4. In countries where corruption is embedded into the political and governmental system, there is an obvious monetary incentive for governments to prevent the production and publication of information evidencing any wrongdoing. While this report was not focused on corruption, the presence of it significantly affects the parliamentary data and information environment, as politicians take a wholesale view that publishing any data might have negative consequences that can easily be avoided by publishing nothing. The personal feelings and attitudes of politicians and parliamentary administrators towards information and data was only touched upon in this study, as access to parliamentarians was understandably difficult to arrange; however, further study into what solutions and incentives might shift attitudes towards data would have significant value for digital technologists and open government advocates. Much data on parliamentary procedure could theoretically be produced that would have little harmful effect on corrupt politicians, and understanding how to move the dial on information production would enable citizens and civil society to have a greater understanding and involvement in governance.

Further research in each of these areas would supplement the findings of this study and enable higher quality decision-making in funding and developing digital tools for citizen engagement in governance.

Conclusions

This report has offered a detailed discussion on how the online production and dissemination of parliamentary and political information occurs in Sub-Saharan Africa, and has identified several key areas where citizens, media houses, civil society organisations and parliaments themselves would benefit from targeted support and development from the international civic technology and philanthropy community to achieve greater reach and effectiveness.

The original research question for this study was:

How do national and sub-national conditions in Sub-Saharan Africa affect how communication between citizens and governance institutions are conducted digitally?

This report has shown that there are several key social and environmental factors affecting how citizens and their parliaments communicate digitally, specifically concerning data access and quality, public understanding of institutional functions, political will to produce and publish information on parliamentary activity, political will to open and maintain communication, the capacity to use relevant digital spaces to consume and share information, and the presence of intermediaries (or infomediaries) to repackage and disseminate information in a widely consumable format.

The main research question was broken down into several sub-questions used to further focus the research and explore the factors in detail, summaries of which are shown below:

1 - In what ways are parliaments creating data/information and making it available?

The creation and publication of information and data by parliaments was variable across the four case study countries, and the level of information and data produced correlated with the relative positions of each of the case study countries on the Open Data Barometer and the Corruption Perception Index. In Uganda, where corruption is relatively high, the data produced is of relatively low volume and quality, whereas South Africa, where corruption is relatively low compared to the other case study countries, demonstrated higher data availability. Uganda was the only case study country not to publish Hansard or a similar record of plenary discussion, one of the most basic pieces of information citizens might expect from their parliament.

Kenya's parliamentary website stood out as the most accessible and relevant to citizens wishing to view or engage with parliamentary information, with a wealth of information set out logically for the user. The parliament websites of all case study countries, however, remain difficult to navigate for individuals that are not very familiar with parliamentary procedure. The PDF format is the preferred file type for parliament websites, which reduces the ability of external groups to automate the publication of information in a more citizen-friendly format. Additionally, parliaments were generally slow to publish information, reducing the ability of citizens and civil society to engage with the legislative process while it remained open to consultation or challenge. In general, parliaments in Sub-Saharan Africa are making some effort to publish information digitally, but not yet in a format that is necessarily accessible, useful or consumable to the general public.

2 - What political and official will exists to engage in, and to improve, two-way information flow between citizens and institutions?

Across the four countries, political will to make progress on open governance principles, including improving two-way communication between parliamentarians and citizens, was relatively low. While certain countries are members of the Open Government Partnership, the rhetoric concerning open governance rarely stretched to instigating action or change. Citizens demonstrated an appetite for engaging with their parliamentarians, in particular digitally through social media or other platforms facilitating communication, but those in power, more often than not, showed little appetite for responsiveness.

Comment on social media is still ignored, email submissions to legislative committees still go undiscussed, and requests for legitimate debate on specific issues are still refused. This poses a threat to digital means of communication in governance matters, where citizens are likely to lose faith in the ability of digital solutions to be effective, as parliamentarians refuse to close the feedback loop. New solutions in the future may experience low take-up, because citizens have seen too many times before that such interventions do not achieve anything. Raising political will to engage and close the feedback loop is therefore key to the success of digital communication.

3 - How are citizens consuming, contributing to, and sharing information on politics and parliamentary institutions?

Citizens are generally not gravitating towards parliamentary monitoring websites, or towards the official websites of their parliament. This is not to say that these monitoring sites are not influential. They absolutely are. They perform an integral infomediary function, without which many journalists, broadcasters, civil society and campaign groups could not perform their professional roles. Citizens themselves prefer to acquire digital information on their parliaments through social media, and they particularly prefer information that has been editorialised, visualised in pictures, diagrams and video, and is easily shareable. Large data files are shunned due to the cost of downloads, and so content that is 'light' in terms of data usage is preferred, and such considerations explain why individuals would rarely download a PDF file from an official parliamentary website. Citizens are also overwhelmingly using mobile technology to consume and share information, and so information made for mobile consumption becomes more popular and widely shared than dense text or websites that are not mobile optimised. The important point to note is that citizens are consuming this information, that there is an appetite for well-packed information, and that it is difficult to quantify with any certainty how far the information cascading down from parliamentary monitoring sites travels.

4 - How does the accessibility, connectivity and quality of the data landscape shape how individuals and institutions communicate?

This report has shown very clearly that the accessibility and quality of data is key in building and maintaining digital tools that enable citizens to acquire and consume parliamentary information. The absence of good data fundamentally means the absence of good automated digital governance tools, with costly and laborious workarounds necessary to fill the gap. While these workarounds evidence heroic efforts on the part of organisations such as PMG in South Africa and ParliamentWatch in Uganda, they rely on significant human resource and the funds to support that activity. Deficiencies in the data environment have shown that there is no lack of appetite for information on what parliament is doing amongst citizens; however, when online communications from citizens are ignored, this frustration spills out into the real world in the form of protests and collective action. This was evidenced in Nigeria and in South Africa at the time of the field work, where individuals tired of the lack of digital responsiveness chose to demonstrate more tangibly.

5 - What other as yet unidentified factors are shaping digital information flow and communication between citizens and parliamentary institutions?

Several factors shaping how parliaments and citizens interact were hypothesised in the earlier sections of this report, and included considerations of social media, political context and data quality. Factors that were not foreseen, and that were uncovered in the course of this research, include the way in which parliamentary systems are structured administratively and electorally, the way in which citizens use mobile technology and data packages to access content and communicate, the use of WhatsApp for communication and organisation on a wider scale, and the use of traditional media to create an amplifier effect for digital tools. Knowing how these previously unforeseen factors shape the landscape of digital governance will enable future digital tools to be built with greater relevance and usefulness to citizens. This knowledge will also enable civil society groups, media groups and civic technologists to understand the possibilities and obligations involved in building and implementing such tools, and the commitment required to maintain them.

What this means

The aim of this research was, as noted above, to understand better the factors that shape how citizens, parliaments and stakeholders communicate official information in Sub-Saharan Africa, but this report should fundamentally be viewed as a call to action on developing digital democracy in the region in a strategic and holistic manner. Large government agencies such as DFID and USAID, philanthropists such as Omidyar Network and the William and Flora Hewlett Foundation, not to mention large international NGOs such as Oxfam, MSF and the UN agencies, invest millions of dollars per year in developing digital forms of democracy in developing countries. The level of failure associated with some of these programmes is completely unknown, as failure is rarely disclosed or quantified by such agencies. This means that mistakes are repeated, and already-stretched resources are wasted. This report shows that success is possible, but that certain conditions must be created or identified as a basis for further development. This report should provide significant food for thought to organisations funding and implementing digital tools for democracy, and if its contents are taken into consideration, should provide future projects with better chances of achieving meaningful impact for real people.

Finally

It is important to note that this report is not exhaustive, and there may be examples of digital tools that were not discussed in this report. The authors covered as much ground and spoke to as many relevant people as possible within the financial and time-limited scope of the project.

In addition to this short summary of the main findings, it is also important to acknowledge the overwhelmingly positive and inspiring digital governance work being conducted by citizens, civil society, media groups and infomediary groups in the case study countries examined. Success in using digital tools to hold parliaments to account looks different in each of the countries examined, and progress towards ultimate success varies; however, the tools in use represent significant steps forward and genuinely demonstrate public benefit, even if that cannot be clearly quantified. While Kenya and South Africa have more advanced digital tools, these have also been operating for much longer and with more stable funding and data than those in Uganda or Nigeria, where digital tools to engage with parliaments are overcoming higher hurdles at present. While social media plays a huge role in sharing information on governance and in citizens engaging publically with parliamentarians, without the parliamentary monitoring sites and other political content-creators such as The Elephant or Africa-Check, suitable and legitimate content would not be available for citizens to share at all. With the right support in funding, and the provision of digital expertise, these sites have the potential to flourish and provide vital information to citizens, as well as facilitate direct contact between citizens and parliamentarians. Whereas it is easy to be sceptical about the overall benefit of digital tools in environments where digital is still a minority facility, these tools can provide benefit beyond visitors to the tools themselves, and ultimately form a key role in the maturation of open and engaged digital government.

Acknowledgements

The authors would like to thank the following organisations for providing help, guidance and support in the research and production of this report:

The authors would also like to thank individuals of the following organisations for generously giving their time and expertise to assist in understanding the issues discussed in this report:

A further two organisations preferred not to be named in the report.

About mySociety

mySociety is an international not-for-profit social enterprise based in the UK and working internationally. We provide technology, research and data that give people the power to get things changed and help them to be active citizens. We work internationally to support partners who use our technology and data in over 40 countries around the world. As one of the first civic technology organisations in the world, we are committed to building the civic technology community and undertaking rigorous research that tests our actions, assumptions and impacts.

Notes


1:World Bank (2018) https://data.worldbank.org/indicator/IT.NET.USER.ZS [accessed 7th August 201]

2:Weymans, W. (2005). Freedom through political representation: Lefort, Gauchet and Rosanvallon on the relationship between state and society. European Journal of Political Theory, 4(3), 263–28.

3:Dahl, R. A., & Flanigan, W. H. (1971). The New Haven Community Study: Yale University, Summer 1959 (No. 7205). Inter-university Consortium for Political Research.

4:Scully, R. & Farrell, D. (2001). Understanding Constituency Representation in the European Parliament. Paper Presented to the Conference of the European Community Studies Association, Madison, May 31 – June 2

5:Scully, R. & Farrell, D. (2001). Understanding Constituency Representation in the European Parliament. Paper Presented to the Conference of the European Community Studies Association, Madison, May 31 – June 2

6:Alcoff, L. (1991). The problem of speaking for others. Cultural critique, (20), 5-32.

7:Goodin, R. E. (2004). Representing diversity. British Journal of Political Science, 34(4), 53–68.

8:Brown, M.B. (2006). Survey article: Citizen panels and the concept of representation. The Journal of Political Philosophy, 14(2), 203–225.

9:Foa, R. S., & Mounk, Y. (2016). The democratic disconnect. Journal of democracy, 27(3), 5-17.

10:Coleman, S., & Blumler, J. (2009) The Internet and Democratic Citizenship: Theory, Practice and Policy, Cambridge University Press

11:Edigheji, O. (2006). Political representation in Africa: towards a conceptual framework. Africa Development, 31(3), 93–119.

12:Oni, S. (2013), The legislature and constituency representation in the Fourth Republic of Nigeria's democratic governance. Ife Psychologia, 21(2), 232-241.

13:Awah, A. (2013). The crucial role of legislation in the promotion of democratic governance vis-a-vis the diversified interests of the electorates and the need for a cordial working relationship between the legislature and other arms of the government. Paper presented at the Pan African Parliamentary Conference on Capacity Building, Rabat, Morocco, 20-22 May.

14:Norris, P., Wynter, T., & Cameron, S. (2018). Corruption and Coercion: The Year in Elections 2017. Electoral Integrity Project annual report. At: https://www.electoralintegrityproject.com

15:Azevedo-Harman, E. (2011). Parliaments in Africa: Representative institutions in the land of the "Big Man". The Journal of Legislative Studies, 17(1), 65– 85

16:Mattes, R. J. D. B. and Mozaffar, S. (2012). African Legislatures: A Glass Half Full? In Ben-Zeev, Keren and Jochen Luckscheiter (eds.) Do Parliaments Matter? African Legislatures and the Advance of Democracy. Cape Town: Heinrich Böll Foundation. pp 8-12

17:Oni, A. A., Oni, S., & Ibietan, J. (2016). ICT and Democratic Parliament in Africa: State of the Matter. Journal of Governance and Development, 12(1), 71-85.

18:Gberevbie, D. E. (2014). Democracy, democratic institutions and good governance in Nigeria. Eastern Africa Social Science Research Review, 30 (1), 133-152.

19:Veit, P., Gracian, Z., Alfred, B., Shamiso, M., Galega, P., Mpundu, K., Rugemeleza, N., Owuor, B., Salomao, A. and Tumushabe, G. (2008). On Whose Behalf? Legislative Representation and the Environment in Africa. World Resources Institute, Washington, DC

20:Pantoja, A. D. & Segura, G. M. (2003). Does ethnicity matter? Descriptive representation in legislatures and political alienation among Latinos. Social Science Quarterly. 84(2), 441-460.

21:Rosenthal, A. (2009). Engines of democracy: politics & policymaking in state legislatures. Washington DC: CQ Press.

22:Oni, A. A., Oni, S., & Ibietan, J. (2016). ICT and Democratic Parliament in Africa: State of the Matter. Journal of Governance and Development, 12(1), 71-85.

23:Bwalya, K. J., Plessis, T. D. & Reinsleigh, C. (2012). Conceptualization of e-parliament in promoting e-democracy: prospects for the SADC region. In Z. S. Mehmet (Ed.) E-parliament and ICT – Based Legislation (pp.295-311): Concept, Experiences and Lessons. Hershey: GI Global

24:Robert, A., & Namusonge, G. S. (2015). The Use of Information and Communication Technology and Social Networking Sites in Political Governance of East African Legislative Assembly Parliament. International Journal of Scientific & Technology Research, 4(6), 136-143.

25:Oni, A. A., Oni, S., & Ibietan, J. (2016). ICT and Democratic Parliament in Africa: State of the Matter. Journal of Governance and Development, 12(1), 71-85.

26:Lengel, L., & Newsom, V. A. (2014). Mutable selves and digital reflexivities: Social media for social change in the Middle East and North Africa. In Symbolic Interaction and New Social Media (pp. 85-119). Emerald Group Publishing Limited.

27:Pickover, M. (2014). Patrimony, power and politics: Selecting, constructing and preserving digital heritage content in South Africa and Africa.

28:Pickover, M. (2014). Patrimony, power and politics: Selecting, constructing and preserving digital heritage content in South Africa and Africa.

29:Bartlett, J., Krasodomski-Jones, A. L. E. X., Daniel, N. E. N. G. A. K., Fisher, A., & Jesperson, S. A. S. H. A. (2015). Social media for election communication and monitoring in Nigeria. Demos, London.

30:Robert, A., & Namusonge, G. S. (2015). The Use of Information and Communication Technology and Social Networking Sites in Political Governance of East African Legislative Assembly Parliament. International Journal of Scientific & Technology Research, 4(6), 136-143.

31:Kamp, M., Messerschmidt, M., & Rugambwa, I. (2016). The Impact of Social Media on The Run-Up To The 2016 Elections In Uganda. REALITY CHECK, 21.

32:Bosch, T. (2017). Twitter activism and youth in South Africa: The case of# RhodesMustFall. Information, Communication & Society, 20(2), 221-232.

33:Rumbul, R. (2016, September). ICT and citizen efficacy: the role of civic technology in facilitating government accountability and citizen confidence. In IFIP World Information Technology Forum (pp. 213-222). Springer, Cham.

34:Anderson, C., Hudson, H., & Feruglio, F. (2018). Appropriating Technology for Accountability: Messages from Making All Voices Count.

35:Anderson, C., Hudson, H., & Feruglio, F. (2018). Appropriating Technology for Accountability: Messages from Making All Voices Count.

36:Roberts, T. and Hernandez, K. (2017) ‘The Techno-Centric Gaze: Incorporating Citizen Participation

37:Font, J., Wojcieszak, M., & Navarro, C. J. (2015). Participation, representation and expertise: Citizen preferences for political decision-making processes. Political Studies, 63(1_suppl), 153-172.

38:Coleman, S., & Blumler, J. (2009) The Internet and Democratic Citizenship: Theory, Practice and Policy, Cambridge University Press

39:IOLNEWS (2018), OPINION: Cashless-society banking requires local solutions, https://www.iol.co.za/business-report/opinion/opinion-cashless-society-banking-requires-local-solutions-17516768

40:Statistica (2017a) - South Africa Social Network Penetration, https://www.statista.com/statistics/284468/south-africa-social-network-penetration/ [accessed 7th August 2018]

41:iFreeKe (2017) - State of the Internet in Kenya Report 2017 , https://www.ifree.co.ke/wp-content/uploads/2018/02/State-of-the-Internet-in-Kenya-report-2017.pdf [accessed 7th August 2018]

42:Daniel, M (2018) - Uganda: why internet and social media stats could all be a lie, https://niinye.avalanch.me/uganda-why-internet-and-social-media-stats-could-all-be-a-lie-96c1ba9693e6 [accessed 7th August 2018]